Professional Documents
Culture Documents
Recording System
®©
Television Broadcasting
Tape and Disc
Recording Systems
by
Harold E. Ennes
Since the first television tape recorder was introduced in 1956, there
has been phenomenal growth in the use of video magnetic tape by telecast-
ers and independent production centers. It is difficult for the "old-timer"
to realize that many of his fellow workers do not remember the days before
tape, when practically two -thirds of the nation watched tv programs at odd
hours or from "hot kines" that had little resemblance to modern recordings.
Video tape, from the very start, provided superior picture quality, giving
a presence that cannot be matched by film. In addition, this medium pro-
vided the telecaster, for the first time, with a capability for local spot -com-
mercial and program production well within the means of a modest budget.
In spite of (or because of) the rapid growth, students and practicing
engineers have been faced with a scarcity of information relative to basic
recording systems. Having been privileged to attend training seminars at
both Ampex and RCA, as well as having had considerable practical experi-
ence with each system, the author has undertaken to fill this need. The pri-
mary purpose of this book is to provide fundamental knowledge for the
practicing engineer who feels the need for a beter understanding of his
equipment; this information may also serve as an introduction to the sub-
ject for the beginner.
Specific circuitry will undoubtedly change in the future as it has in the
past. Therefore, this coverage is general in nature, pointing up the primary
functions of video -tape equipment. The scope of coverage is from basic
theory to testing and maintenance of complete systems.
This text assumes a background in basic electronics theory. The reader
should also have a fundamental grasp of semiconductor logic circuitry, at
least of the same level as that presented in Chapter 3 of Television Broad-
casting: Systems Maintenance by Harold E. Ennes (Indianapolis, Howard
W. Sams & Co., Inc., 1972) . For the serious reader or student of broad-
casting, the following texts (by the same author and publisher) should be
considered prerequisites to this book: Workshop in Solid State, Television
PREFACE
HAROLD E. ENNES
Contents
CHAPTER 1
CHAPTER 2
CHAPTER 3
CHAPTER 4
CHAPTER 5
AND FM SYSTEM 89
5 -1. Functional Description 89
5 -2. Basics of the Modulation Process 92
5 -3. Demodulation Basics 95
5 -4. Refinements in Modern
Modulation -Demodulation Systems 102
5 -5. Modulation Circuitry 106
5 -6. Combined Modulation -Demodulation
Frequency Response 121
5 -7. The Video Preamplifier 125
5 -8. The Playback Amplifier 128
5 -9. Electronic Channel Switching 131
5 -10. Modern Demodulation Circuitry 145
CHAPTER 6
CHAPTER 7
CHAPTER 8
CHAPTER 9
CHAPTER 10
CHAPTER 11
CHAPTER 12
CHAPTER 13
CHAPTER 14
APPENDIX A
APPENDIX B
INDEX 569
CHAPTER 1
Velocity
the system is fixed by head -gap size and tape speed. Response can be ex-
panded into higher frequency regions by either or both of two methods:
(1) increased tape speed to produce a longer wavelength for a given high
frequency, (2) decreased gap size. See Fig. 1 -3.
Note that the higher tape speed extends the upper frequency response,
but, since the wavelength for a given low frequency is longer, the low-
Medium Frequency
Magnetic Gap
Between Pole Pieces
Higher Speed
m
inisi
yl
Fig. -3. Effects of head gap and
1
I-
_r AI
f xs
SmallerIn
aim.
Noise Level
ing the gap size means that less energy is available at the lowest frequen-
cies, so the low- frequency response again suffers.
The strength of the magnetic field is determined by the amplitude of
the signal fed to the head and the rate of change of the magnetic field.
For a given signal amplitude, as the frequency is increased (increased rate
of change) the magnetic -field intensity is increased. As the frequency is
doubled (increased by one octave) , the voltage is doubled. This voltage
increase, measured in decibels, is 6 dB. Therefore, a 6 -dB -per- octave rise
with increasing frequency occurs in magnetic recording. Fig. 1 -4A shows
the typical response (6 dB /octave) up to the high- frequency limitations,
where the response falls quite rapidly. Note that near the noise level and
below, the response cannot be called usable. The noise levels are tape
noise, modulation noise, and noise level of input amplifiers.
Conversely, as the frequency decreases from a given frequency limit, a
6 -dB -per- octave falloff in response occurs. At a certain low- frequency
limit, the change rate of the magnetic recording is so low, compared to the
gap size, that very little output exists. When the signal output falls to an
objectionable signal -to -noise ratio, the lower frequency limitations of the
system have been exceeded.
Fig. 1 -4B shows the typical equalization curve for correction of the mag-
netic recording characteristic of Fig. 1 -4A. The rising response at the end is
6 dB/Octave
CC z
to correct for the rolloff at highest frequencies in the passband due to gap
size and tape magnetic- coating factors.
The frequency limitation of direct magnetic recording is about nine
octaves, or 30 to 15,000 Hz in the audio range. If it is desired to increase
the upper frequency limit to 30,000 Hz, the lower frequency limit must
also be doubled to maintain an adequate signal-to -noise ratio. The result is
still a nine -octave range, 60 to 30,000 Hz.
The vertical resolution determines how well the horizontal lines in a pic-
ture are resolved. The maximum vertical resolution is fixed by the number
of active scanning lines. The United States standards call for a total of 525
lines per frame. Vertical- blanking time is approximately 7.5 percent of the
total frame time; therefore, the number of lines blanked out is 525 X 0.075
= 40 (approx) Since approximately 40 lines of the picture are blanked
.
out, approximately 485 active picture lines remain. This would appear to
indicate that 485 vertically spaced horizontal lines would be resolved, but
in practice this is not true. The slight spacing between the scanning lines
and the fact that the scanning spot straddles some of the lines both tend to
reduce the utilization of the maximum number of active lines. The reduc-
tion factor is usually considered to be 0.7 times the total active lines. Thus,
there are 485 X 0.7, or approximately 340 usable lines.
Fig. 1 -5 shows the horizontal wedges of a test pattern. Here, the lines
merge at a point that represents 340 black and white horizontal lines in the
total image height. This is a typical value of vertical resolution at both the
BASIC CONCEPTS IN QUADRUPLEX VIDEO TAPE SYSTEMS 17
studio and transmitter outputs; hence, the television tape system should not
limit it.
Horizontal resolution is the ability to define vertical lines in the image.
The essentially round shape of the scanning spot and the fact that it is not
infinitely small place an immediate limitation on the ability to reproduce
rapid picture transitions. Thus, when the beam suddenly encounters the
sharp vertical line representing a transition from black to white ( Fig. 1 -6) ,
the resulting signal is not a square wave, but more nearly a sine wave.
Scanning Beam
--o- White
Black
Direction of Scan
Another factor contributing to the sine wave is the fact that a straight
vertical line represents an infinitely short rise time, which would require
infinite bandwidth. Such a bandwidth, of course, is impossible in practice;
the rise time of the signal representing the instantaneous transition is
limited by the available bandwidth of the system.
When the total of the rise and decay time equals the spacing between
lines, the lines are not visible as separate picture elements and are not
resolved. Rather, they appear as a solid area (Fig. 1 -7) . Assuming that the
scanning beam is properly focused, the limitation on horizontal resolution
is the system bandwidth. The pulse rise time representing an instantaneous
transition in the picture is directly related to the system bandwidth. As a
"rule of thumb," resolution of 80 tv lines requires a 1 -MHz bandwidth.
The gain-bandwidth product defines the relationship between a practical
amount of gain and the necessary bandwidth:
f = Gain X Bandwidth
where,
f is the upper frequency limit for a gain of unity.
For example, for a given circuit with gain reduced to unity at 35 MHz:
Gain X Bandwidth = 35 MHz
Then for an uncompensated amplifier, the bandwidth for a gain of 10 is:
35 MHz
Bandwidth =
10
= 3.5 MHz
The relationship of bandwidth and rise time is of equal importance. The
product of bandwidth and rise time gives a factor, k, which depends on
type and magnitude of high- frequency compensation. If leading -edge over-
shoot is limited to less than 3 percent, factor k can be given a value of 0.35,
and the relationship between bandwidth and rise time can be expressed in
any of these three forms:
BW X RT = 0.35
0.35
BW =
RT
0.35
RT =
BW
Thus, the greater the bandwidth, the shorter is the rise time that can be
passed. For a bandwidth of 10 MHz:
RT =05
10
= 0.035 µs
1 0.35 80
2 0.175 160
3 0.1166 240
4 0.0875 320
5 0.07 400
6 0.058 480
7 0.05 560
8 0.0437 640
9 0.039 720
10 0.035 800
The video- signal frequency range for television tape extends from very
low frequencies (actually dc) to an upper frequency of at least 4 MHz. If
the range is considered to be 10 Hz to 4 MHz, a gamut of more than 18
octaves is required, which, as discussed earlier, is not possible in direct
magnetic -recording techniques.
recording systems.
white deviates 1.8 MHz upward in frequency; the peak -white signal occurs
at 6.8 MHz in Fig. 1 -8. With a standard video input, 0.3 -volt sync to 0.7-
volt video, sync tips deviate the carrier 0.7 MHz lower in frequency, to
4.3 MHz. The total deviation of 2.5 MHz is used currently as the 100 -per-
cent modulation reference for monochrome signals, in low-band video
tape systems.
When the modulation index is less than 0.5 ( frequency deviation less
than one -half the modulation frequency) , sideband energies beyond a single
pair are practically nonexistent, and the modulated signal approaches a -m
characteristics in this respect.1 Sidebands occur at the carrier frequency plus
At 2.5 MHz
/
\
1 MHz 4.3 MHz 5 MHz 6.8 MHz
/ \ /I 1
\\
I
Video
Fig. 1 -8. Frequency modulation of carrier for low -band monochrome recording.
and minus the instantaneous video frequency. The maximum video fre-
quency in the system passband determines the sideband limits. For a 4 -MHz
-
bandwidth, the lowest sideband occurs at 5 4, or 1 MHz. The upper side-
band extends just far enough to provide a shelf for the upper frequency
deviation. Sidebands beyond this limit are not used in low -band recording.
The only practical result of these compromises is a slight zig-zagging of
vertical lines apparent at a resolution of 300 lines and above.
The total frequency range now becomes 1 MHz to approximately 7 MHz;
thus, the modulation process has reduced the original 18- octave video range
to less than three octaves. A practical solution to magnetically recording
the video signal has been found if the problem of handling a 7 -MHz signal
can be met.
A good audio tape recorder may have a head -gap size as small as 0.25 mil
(0.00025 inch) Then, for an upper audio limit of 15,000 Hz and a tape
.
Record-Playback Relay
Commutator
Ground
II - II
Four Video Heads on
Drum Rotating At
14,400 rpm 1240 rpsl
I I
I
I
dl
Mono I q`I
Ir
I3
I
c`I
I
I Mono I
m1
ml
5.5 5.79
I Id
I 6.5I
{ I
4. 28 5.0 I I 16.8 7. 06 7.9 10.0
i
I I i MHz
I
MHz
I I I I
Color Color (
I '
1
At1 MHz F- H
Wavelength - Frequency
Velocity
then,
50% Gray
1 Volt p-p
Picture Black 7.5% Setup
_ Blanking
Sync Tip
Applied Video
I
Fig. 1 -11. High -band recording standard for monochrome and color signals.
BASIC CONCEPTS IN QUADRUPLEX VIDEO TAPE SYSTEMS 23
to record one hour of material. Try to visualize the required tape -transport
mechanism; obviously, this is an impractical solution.
The problem is solved by pulling the tape at a practical speed of 15 in /s
past a rotating video head. Insofar as the video signal is concerned, the
resultant velocity is more accurately termed head -to-tape velocity. The
rotating -head principle, illustrated in Fig. 1 -9, gives an effective head -to-
tape velocity of slightly over 1500 in /s. The fm video signal is fed to the
individual heads (quadruplex system) on the rotating drum and brush
assembly. Video tracks approximately 10 mils wide are laid down trans-
versely across the 2- inch -wide magnetic tape. Audio is recorded longi-
tudinally along the top of the tape at the conventional 15 -in /s rate. A 240 -
Hz control signal, which indicates the precise position of the rotating head
drum relative to the recorded video information, is recorded longitudinally
along the bottom of the tape. (NOTE: Details of connections between the
heads and the commutator are given in Chapter 2.)
VideolAudio
Master System Tape Waveform Control (Also
Erase Control Transport Video & Audio Monitor Picture Waveform Electronic Editor
Head Panel Panel Head Cover IA Scope) Monitor Monitor and Editec Panel
Modern television tape recorders of both Ampex and RCA design are
housed in a single console requiring no external rack space. External racks
normally house video and audio signal- source selection panels, color moni-
tors, and vectorscopes.
A typical tape "system" is the Ampex VR -2000 video tape recorder
shown in Fig. 1 -12. The rotating video head drum and all audio heads are
under the protective head cover. The front of the lower section houses all
system modules, and the rear of the lower section contains the required air
systems and intermodule wiring harness.
Monitor
Deck
Secondary
Control Panels
(Cover Open)
Transport
(Cover Open
Primary
¡- Control Panel
Normal system inputs are ac power, selected video and audio signal
sources, composite sync, and 3.58 -MHz subcarrier.
Fig. 1 -13 illustrates the Ampex AVR -1 tape system. This is a complete
and automatic system itself, or it may be used on a "shared electronics"
basis for the Ampex ACR -25 video cassette system (Chapter 13) .
26 TELEVISION BROADCASTING
EXERCISES
Q1 -1. From your studies thus far, and your understanding of Figs. 1 -3 and 1 -4,
would you conclude that the low video frequencies need to be brought
up in amplitude relative to higher video frequencies?
Q1 -2. If you have an amplifier with unity gain at 50 MHz, what is the usable
bandwidth at a gain of 5 (uncompensated) ?
Q1 -3. Assume that the head -to -tape velocity is 1560 inches per second. What
is the recorded wavelength of a 10 -MHz signal?
Q1 -4. What is the shortest usable wavelength on the tape for a head gap of
50 microinches?
Q1 -5. What frequency corresponds to your answer for Q1 -4, assuming the
same velocity as in Q1 -3?
Q1 -6. What is the term applied to the sideband above the highest deviation
frequency in video tape recording?
Q1 -7. In the drawing of Fig. 1 -8, a standard 1 -volt p -p composite signal input
is assumed. Give the sync -tip frequency and peak -white frequency re-
sulting from an input which overshoots to 1.5 volts p -p (video -to -sync
ratio maintained and no video clipping or limiting used in recorder).
CHAPTER 2
You will now learn the precision of the four-head (quadruplex) assem-
bly. By acquiring the time -base concept, you will more clearly understand
the relationship of the recording system to the television signal itself.
You will learn later that as precise as the rotating head assembly is, more
than one -half of all the rest of the system function is concerned with cor-
recting for deficiencies of this assembly. This is the reason why the time -
base concept is so important.
The circular mounting which contains the four video heads is termed the
headwheel by RCA, and the drum by Ampex.
Start
then 0 0.54°
Track Slant
End
Tape Travel at 15 in /s
AHeadwheel Video
Tape
Travel Sound Track - Bottom of Taoe
F- Control Track
145 Mils)
Top of Tape
Top View
Tape Travel
Cue Trac k
70 Mils- [-
1.82-
Looking Toward Tape
From Headwheel Side 2"
Sound
Tape Travel Track Top
Air
/,
'
'
sure
Vacuum Guid
r
A
Head
Tip
Penetration
V
.4
Tape
orientation on the tape is that which is viewed looking toward the coated
side from the rotating heads. The track dimensions on the horizontal
version shown in Fig. 2 -2B are exactly the same as for the vertical arrange-
ment in Fig. 2 -2A.
The method of holding the tape concentric with the rotating heads is
shown in Fig. 2 -3A. The vacuum guide holds the tape in place, and the
center slot provides clearance for tape stretch under head penetration.
Fig. 2 -3B is a side view of the vacuum guide and the rotating headwheel.
Most modern tape systems provide for a dual -speed tape transport (71/z
or 15 in /s). Dual -speed operation employs standard quadruplex recording
techniques, using standard two -inch recording tape. All the advantages of
quadruplex recording are retained, and the cue track, audio track, and
control track are untouched. Only the method by which video tracks are
laid down is modified.
The headwheel employs four video heads which record a track only 5
mils wide, as compared with 10 mils for conventional recording. Fig. 2 -4A
shows how the 5 -mil tracks at half speed are spaced with respect to the
2.5 -Mil
1111111111111111111111111111111
10 -Mil 5-Mil
Spacing 1111111111111111111111111111111 Spacing Spacing
video portion of the tape (5 -mil track with 21/2 -mil spacing) . The capstan
motor pulls the tape at 71/2 in /s; thus, twice as much video information is
packed into each foot of tape. Because the speed of the headwheel rotation
is the same as in normal operation, each recorded track contains the same
number of tv lines, approximately 16.5.
During operation at 15 in /s, video tracks are recorded by the half -track
heads as shown in Fig. 2 -4B. In this case, the track width remains at 5 mils,
but the spacing between tracks becomes approximately 10 mils. The tracks
in Fig. 2 -4B correspond to every other track in Fig. 2 -4A.
Standard video -headwheel panels recording 10 -mile tracks with approx-
imately 5 -mil spacing between them (Fig. 2 -4C) are oriented in the same
manner as in Fig. 2 -4B. Comparison of Figs. 2 -4B and 2 -4C will show how
interchangeability is achieved.
If a 10 -mil head traces a 5 -mil track, the head is in contact with the
signal included on the track, and, in addition, it traces over 5 mils of blank
tape. It encounters the next track after a 5 -mil spacing interval which
brings the head into position to read all the information on the next track.
In the case of a half -track (5 -mil) head reading a fully recorded (10-
mil) track, the principle is reversed. Here, the half -track head reads enough
of the video information (5 mils) on the fully recorded track to produce
high -quality pictures.
A B C
I ¡
Recording Circumference
The recording diameter at the minimum tip projection of 1.0 mil is:
A +B +C =0.001
2.064
0.001
2.066 inches
The recording circumference is:
Circumference = and
= 3.1416X 2.066
= 6.490 inches
The head-to -tape velocity is:
6.490 X 240 1557 in /s
at minimum tip projection of 1.0 mil.
That is, the space occupied by a tv line must be the same at the middle and
end of the video track as it is at the beginning of the track.
Fig. 2 -6A shows the tape without the head engaged. As the vacuum
guide is moved toward the head, the pole pieces contact the tape and a
stretch occurs as shown in Fig. 2 -6B. The greater the tip penetration, the
greater is the stretch and the wider is the space between sync pulses. The
effective spacing in time between pulses is determined by the tip velocity.
A recording made with a 3 -mil penetration can be played back with a worn
head having a tip penetration of only 1 mil. This is because the reduced
tip velocity of the worn head means a longer time is required to scan a
given length of the video track; therefore less tape stretch is required to
maintain the correct space-time relationship between pulses.
Pulses
Farther
Apart
Equally Spaced
Sync Pulses
(A) Tape without head engaged. (B) Tape with head engaged.
Fig. 2 -6. Pole -tip stretch of tape.
From this analysis, it is evident that reduced velocity due to worn tips
is complementary to the resulting reduction in tape stretch. Therefore,
assuming that the video -tape stock neither shrinks nor stretches during its
life, the vacuum -guide position relative to the headwheel remains the same
as tip projection decreases with head wear.
To emphasize this point, see Fig. 2 -7. A new head with a 3.6 -mil tip
stretches the tape considerably more than when this same head is worn
down to a 1 -mil tip. But note that the head -to -tape velocity is 4 in /s
greater for a 3.6 -mil projection than when the head is worn to a 1 -mil
projection.
This is not a geometric (space -in-time) error if the vacuum guide has
not been moved, since the standard radius of rotation is maintained. In
Fig. 2 -7B, the arrow corresponding to maximum tape stretch represents
the length of the pole -tip contact with the tape, which is about 16 video
lines. The other arrow indicates the shorter length of the same 16 lines
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 33
under minimum stretch, but the time taken to scan the same space is
longer; hence there is no geometric error.
Keep the time -space concept. Errors in timing ( velocity errors) result
in horizontal displacement of vertical lines in the picture (Section 2 -5) .
Velocity errors result when the time to scan a given space (such as the
space occupied by 16 lines) differs from the timt elapsed in recording
that space on the tape.
Radius of Rotation Between 1.0329 and 1.0356 Inches
Effective Diameter 2.0658 to 2.0712 Inches
Difference in Stretch
Pole Tip @ 1 Mi
Pole Tip @ 3.6 Mil
Drum
Minimum St etch 6
Length Maximum Stretch
1 Length
@ @ ' ,
1561Ints 1557 In /s
Fig. 2 -7. Change in tip projection does not cause: geometric error.
J
A nonstandard recording radius can be used to -flake a recording on a
tape. If the tape is played back on the same head with the same adjustment,
no geometric error results. This simply means that the playback space-in-
time velocity is the same as the recorded signal space -in -time velocity.
For interchangeability of tapes and to facilitate splicing of sections from
different types, it is necessary to adhere to a standard pole -tip arc radius,
within very narrow limits. Even when this standard Tolerance is maintained,
velocity errors can result from slightly different guide-height adjustments,
tolerances in tape tensions between machines, slightiy different positions of
tape-transport fixtures, and tape stretch resulting from use. (These prob-
lems are expanded in Section 2-5.)
The position of the vacuum guide, which cups the tape around the
rotating heads, determines the amount of pole -tip penetration into the
tape. When the guide is sufficiently far away from the pole tips, a positive
34 TELEVISION BROADCASTING
clearance between tape and heads occurs, and no contact exists. Under this
condition, no recording would be laid down on the tape, and no playback
of recorded information can occur (Fig. 2 -8A) .
An adequate magnetic contact must prevail at the low values of tip pene-
tration which exist near the end of head life. Therefore, since the guide
position relative to the head is a constant value (Section 2 -2) , new heads
with high tip projection will exhibit relatively high values of tip
penetration.
High values of tip penetration increase head and tape wear and the load
on the head-motor drive amplifiers. If excessive loading occurs, head -servo
instability can occur. This is particularly true when marginal tubes are in
service in the older tube -type recorders.
A tip penetration of less than 1 mil aggravates dropouts and possible
head clogging. Dropouts are white flashes in the reproduced picture; they
are caused by microscopic irregularities in the magnetic coating of the tape.
Head clogging results when iron -oxide particles, loosened from the tape
coating, become lodged in the head gap. (See Chapter 8 for tape-dropout
compensation systems.)
2 -4. TIMING
The headwheel (or drum) rotates at 240 rps, or four times the tv -field
frequency of 60 Hz. Thus, one revolution (360°) of the wheel lays down
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 35
four tracks (four heads spaced 90° and fed simultaneously) which are
equal to one -fourth of a field. Since there are 262.". lines per field, there
are 65.625 lines in four tracks on the tape, and each track contains 16.4
lines.
NOTE: A tv line is designated as H, where H is the interval from the
leading edge of one horizontal -sync pulse to the leading edge of the next
horizontal -sync pulse. (The standard interval is 63.5 microseconds.)
Therefore, a recorded track equals 16.4H.
= 4167 microseconds
Time for 90° =4 4 = 1042 microseconds
Since the duration of one line is 63.5 microseconds, the number of lines
in 90° of rotation is:
1042
63.5
- 16.4 lines
Then the number of degrees per line is:
90 = 5.48°
16.4
and the number of lines per degree is:
1 = 0.182 line
5.48
The heads are 90° apart and lay down 16.4 tv lines in 90° of rotation.
Since the heads are spaced 90 °, the remaining time that both heads are in
contact with the tape is 102 ° -90 °, or 12 °. This leaves an overlap of
0.182 X 12, or 2.2 lines (Fig. 2 -9) In recording, the same information is
.
Headwheel
Rela- Degrees
Live Vertical 360° =
Fre- Micro- (V = 1/60 TV Lines 1/240 sec
quency seconds sec = (H = = 4166.6
Subject (Hz) (µs) 16,667 As) 63.5 As) As
-
(240 4) recorded on 15 inches of
a field. Therefore, there are 60 fields
tape. Thus one field equals 0.25 inch (15 _ 60) , and a frame (2 fields)
equals 0.5 inch of tape.
The foregoing timing relationships and other significant factors are
summarized in Table 2 -1.
NOTE: The tracks on video tape can be made visible by coating the tape
with a suspension of carbonyl iron and diluent (Fig. 2 -10) Horizontal - .
sync pulses show as white dots between the long vrhite lines of vertical
blanking. Fig. 2 -11 shows how the picture raster is made up of consecu-
tive bands of 16 to 17 lines contained in each head pass across the tape.
Visible tracks such as those in Fig. 2 -10 are used only for mechanical
splicing. This method has largely been replaced ly electronic splicing
and editing systems (Chapter 10) .
White marks in this control -track area are frame edit pulses.
112"
1 Frame
a'
1 Field
I
2
4
1
Start of Vert
Blanking and Sync
recorded standard tape will yield a picture that, when viewed without any
automatic time correction (ATC) or other time -base correction (Chapter
6) , is free of geometric errors.
If the guide is now moved away from the headwheel as in Fig. 2 -13B, we
have the condition already discussed relative to Fig. 2 -8B. A high head -to-
tape pressure occurs at the top and bottom of the head pass, and a low
I
A
Vacuum Guide
(Moves In and Out)
High
Head -to -Tape
Pressure
//
Radius of Pale
,Tip Rotati:n
\ \
l'l
Low
Head to Tape
Pressure
- i--I
I
Original
Concentric
Center
--'y- Fit
Moved Away
Center of
Relating Wheel
1
'
I
I
\
1\\ /
High
Head -to -Tape
Pressure \\ ././///
pressure occurs at the center. Note that a measure of actual tip projection
of the headwheel is that obtained when the contact at the center of the tape
is lost under nonconcentric (eccentric) operation. We may now study the
various forms of geometric distortion, starting with the example just cited,
the vacuum guide adjusted too far from the headwheel.
Tape Vacuum Guide Too Far From Head
Point Z in Fig. 2 -14A represents the center of the drum, or headwheel.
Assume that the recording is made with arc X -X representing 16 lines.
Y
Enters Late
X -X - 16 Lines on Tape
Y -Y 16 Lines on Tape
Y(
lllllllllllllllllll
/
Lagging Skew
/ / /
/ / / / / / / / / / / / / / /l
ps Error
I
i
'llllllllllllllllll
/ / /
/ / / / /
lllllllllllllllll
/ / / / / / / / // /
Arc Y-Y is the playback position, if the guide is too far from the heads,
and contains the same 16 lines of information. The pole tip enters this
information late and leaves early compared to the recorded time base.
Space errors resulting from the condition in the preceding paragraph
are depicted in Fig. 2 -14B. Each band of lines representing a head pass
across the tape starts late and ends early, producing the skew effect, also
called the venetian-blind effect.
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 41
Interchange S- Distortion
Note from Figs. 2 -14B and 2 -14C that the skewed lines are not straight
diagonal lines, but that each segment is a 90° section of a sinusoid. A
single head scan across the tape is 90° of the headwheel rotation rather
than the full 360 °, so only 90° of the sinusoid occurs on each band of 16
to 17 lines.
Note also from Fig. 2 -14B that the number of microseconds of error
(peak-to -peak) on vertical lines in the picture is approximately in direct
ratio to the guide error (tip -projection error) in mils. Thus if the guide is
misadjusted by 0.5 mil, a timing error of approximately 0.5 µs occurs.
If the guide is misadjusted 1 mil, an error of approximately 1 µs results.
This error is approximate because a 1 -mil misad¡ustment of the guide
results in exactly 0.92 Its of timing error rather than 1 µs, but the rule of
thumb is to use a direct microsecond -to -mil relationship. This information
is useful in judging the amount of tips remaining on a headwheel, as
outlined in Chapter 11.
Note from Fig. 2 -12 that the tolerance on the guide radius is from
1.0334 inches maximum to 1.0329 inches minimurr. Thus the tolerance is
-
1.0334 1.0329, or 0.0005 inch (0.5 mil) Let us see what this seemingly
.
Now if this guide is backed away from the headwheel assembly 0.5 mil,
concentric operation is lost (Fig. 2 -13B) . If the recording had been made
with this same headwheel panel, the result would be as in Fig. 2 -15B, which
is the same as was indicated in Fig. 2 -14B. In this case, however, the guide
is actually being adjusted to obtain minimum skewing of the recorded
signal. Since the recorded signal was made with a guide radius 0.5 mil
larger, the playback guide radius will be made 0.5 mil larger (0.5 mil
lighter pressure) by backing the guide away 0.5 mil from precisely con-
centric operation.
0.5us -¡
A Signal
B Signal
II
20ns-1
See Fig. 2 -15C. Note that the cancellation of timing error is not perfect,
due to the curvature caused by nonconcentric operation of the panel. As
shown by Fig. 2 -15D, the resultant "interchange S- distortion" is about 20
nanoseconds peak -to -peak at the tolerance extremes. Although the stated
guide dimensions are reasonable manufacturing tolerances (0.5 mil, or 500
microinches, is about 1/6 the thickness of this page) , manufacturers are
holding even closer tolerances. Furthermore, such errors are readily reduced
to practically zero by automatic time correction circuitry, as discussed in
Chapter 6.
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 43
Enters Early
X-X = 16 Lines
Y -Y = 16 Lines
X -X -Z = Recording Dimension
)Z
Y -Y -Z = Playback Dimension
- - --
(A) Recording and playback dimen rions.
- - Leaves Late
Starts Early
Ends Late
\ 16 Lines
16 Lines
1
16 Lines
Leading Skew
of the tape with the pole -tip circumference. A major problem in the inter-
changeability and interspliceability of the tape is maintaining the same
stretch of tape at top and bottom of the pole -tip arc.
1. With the guide too high (Fig. 2 -17A), less stretch occurs at the top
than at the bottom; hence, there is lack of concentricity.
2. Assuming the tape was recorded with correct concentricity, the pole
tip enters late and leaves late compared to the recording time base
(Fig. 2 -17B).
3. The start and end of each band of lines is late, producing a scallop, as
shown in Fig. 2 -17C.
4. Vertical lines on a video monitor appear as shown in Fig. 2 -17D
when the guide is too high.
If a recording is made with the guide positioned as shown in Fig. 2 -17A,
it can be played back on the same head without distortion. However, if
Enters Late
Y'
Starts Late
Ends Late
Scallop
this tape is played back on a correctly adjusted head assembly, the guide
must be misadjusted vertically to eliminate scalloping in the picture. A
misadjustment such as that in Fig. 2 -17A produces excessive wear on the
bottom half of the tape.
Tape Vacuum Guide Too Low
1. When the guide is positioned too low (Fig. 2 -18A), more stretch
occurs at the top than at the bottom of the tape.
Leaves Early
Starts Early
Ends Early
Scallop
3. Each band of lines starts and leaves early (Fig. 2 -18C), producing a
scallop which is opposite to that of Fig. 2 -17C.
46 TELEVISION BROADCASTING
f t
Head 2 Displaced
by O Degrees
(A) Head.
Normal Normal
Starts Late -I 2
111111111111111
Fig. 2 -19. Space distortion, head -quadrature error.
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 47
Gap 1
3 ,
(A) Head location. (B) Tracks on tape.
Fig. 2 -21. Axial- position error.
Fig. 2 -22 illustrates the Ampex Mark Ten head assembly. The complete
assembly is built on a rigid base plate, and the base plate is fastened to the
top plate of the transport by one knurled and two hex-head screws. Two
integral connectors on the underside of the base plate engage mating con-
nectors integral with the top plate.
The servo -controlled, hysteresis- synchronous, air-bearing drum motor is
driven by square-wave (instead of sine -wave) ac power, generated by the
Intersync servo system, which also controls its instantaneous frequency. The
nominal frequency is 240 Hz.
The drum -motor armature assembly is supported by air bearings which
cause it to float in a nearly frictionless condition on a thin layer of air.
The air used is furnished at 55 psi (nominal) through an opening (on
the underside of the base plate) which mates with an air line from the
bearing air supply. Due to the dependence of the armature upon air under
pressure for its support, it is imperative that it never be turned or rotated
in the absence of the bearing air pressure. To do so may seriously damage
the air- bearing surfaces.
The video head drum is supported by and mounted on the armature
shaft of the drum motor. The four video heads are mounted on the head
drum in precise quadrature. The drum motor turns the video head drum
at 14,400 rpm.
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 49
Vacuum Hose
Channel l
H
Channel 2 -a-
Fig. 2-23. Ampex head -
identification method.
Channel 4
Channel 3
The nuvistor plug -in preamplifiers are built on one etched board
adjacent to the rotary transformer. Each is a one -stage amplifier which
functions during the system play mode and is biased off during the system
record mode.
The connector at the right end of the video head assembly (Fig. 2 -22)
carries the outputs of the individual nuvistor preamplifiers, as well as their
incoming power. This connector mates with another at the end of a cable,
which routes the preamplifier outputs to their associated preamplifiers in
the head channel housing, and also connects the power required by the
nuvistors.
The surfaces of the vacuum tape guide which face the video head drum
are curved to the circular path described by the rotating video head tips.
Vacuum is applied to a narrow cavity between these surfaces to hold and
cup the tape on this curve. A shoulder on the lower end of both curved sur-
faces stabilizes the position of the lower edge of the magnetic tape.
The vacuum tape guide is hinged to allow it to swing away from or
toward the rotating headwheel. Fig. 2 -24 shows the vacuum guide swung
back from the head on the RCA headwheel panel. When the guide is
closest to the head drum (Fig. 2-22 ) , the guide latch acts to hold it there;
when it is desired to open the guide to gain access to the video heads, the
guide latch is actuated to unlock the guide from the closed position.
On the Ampex assembly, there are two factory -set controls of the position
of the vacuum tape guide. The first is the guide- height knob; the second is
the REC /PLAY lever. When the guide is closed and the REC /PLAY lever is
set at the record position, the vacuum tape guide accurately positions the
tape for standard tip penetrations of the tape.
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 51
Tonewheel Headwheel Motor Guide Guide Height Slip Ring and Reference
(Tachometer) Motor Air Shield Release Adj Knob Brush Cover Button
Lever (Scallop)
VOA bff
the guide and disengage the tape from the video head tips.
The RCA headwheel panel Assembly ( Fig. 2 -21) plugs into receptacles
on a recessed cast mounting on the tape -transport panel. Three captive
thumbscrews secure the headwheel panel to the :ape- transport panel and
facilitate removal and replacement. The headwheel motor rotates on an air
bearing to drive the video headwheel. A jumpe- wire connected to the
headwheel panel activates the high -pressure air system when the headwheel
panel is plugged into position on the tape-transport panel. Removal of the
headwheel panel stops the high - pressure air system from operating.
NOTE: The headwheel panel is a precision instrument that requires
special factory procedures using special tools and instruments for main-
tenance. Therefore it is strongly recommended that parts replacement be
avoided and maintenance be limited to the procedures described in the in-
52 TELEVISION BROADCASTING
The headwheel motor, headwheel, slip rings, and brushes form a single
assembly. The three -phase asynchronous motor rotates at 240 rps. It is
driven by three power amplifiers (one for each phase) in the headwheel
servo system. Fig. 2 -25 shows the physical appearance of the Ampex Mark
XX video head. This contains a special 6 -phase motor winding instead
of the conventional 3 -phase type. It is used in the Ampex AVR -1 tape
system, and the ACR -25 video cassette system (described in Chapter 13.)
The motor is surrounded by a metal shroud which serves both as a mag-
netic shield and an air duct. Cooling air is drawn in at the headwheel end
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 53
of the motor, passes over the motor housing, and i... exhausted to the head-
wheel blower through a hole in the headwheel mounting plate.
Fig. 2 -26 illustrates the RCA method of identifying the video heads.
Compare this to the Ampex method of Fig. 2 -23. Nine silver slip rings
mounted on the headwheel shaft serve to connect the video -head signals to
associated brushes and output connector pins. Each of the four heads has
separate input and return wires to isolate their respective signals and
r Rec/Play Ch 1
Changeover Relay Playback
Ch 1 _04> o t Ch 2
Ch 2 _o_>> o t Ch 3
Ch Ch 4
}
Rec Amp Playback
Ch 4 Playback
Rec Amn Brush
Channels
Returns
`
R Head
Cec urrent
Slip Rings
and
Brushes
r 9
Motor
' Headwheel
Nearest to Motor
Rotor
Head 4
Gd
Headwheel Panel
reduce cross talk. The rotor ground connection s made with a final con- :
nection to the ninth slip ring. Two carbon brushes associated with each
slip ring are spring loaded in pairs to equalize forces and increase reliability.
Note that the heads on the rotating wheel are numbered in sequence,
1- 2 -3 -4. However, head 1 is connected to the first pair of brushes (1 and
2 ) , while head 2 is connected to pair 5 and 6. Noie that if the RCA method
54 TELEVISION BROADCASTING
guide arm (and video tape) perpendicular with respect to the headwheel
panel. This provides the scallop adjustment required during setup pro-
cedures.
After the vacuum guide has been locked in the general operating area
with the vacuum -guide release lever, tape -penetration pressure is provided
by the guide -arm assembly. The guide -arm assemDly is spring -loaded to
press an eccentric bearing against the edge to the vacuum guide and move
the guide and tape away from the rotating headwheel when air pressure is
removed. This is a safety feature that prevents the headwheel from dam-
aging the tape when the machine is stopped, or during other operations
which require the tape to stop while the headwheel continues to rotate.
The guide -arm assembly consists of a round shalt with an arm clamped
at one end and a bearing mounted eccentrically at :he opposite end. A flat
head at the upper end of the shaft contains an extension of the shaft near
the end of the flat head. The bearing is mounted on this extension to pro-
vide an eccentric pressure to the edge of the vacuum guide. Since the
vacuum -guide release mechanism provides a spring pressure pulling the
vacuum guide closed, the eccentric gear holds the vacuum guide open to
the required penetration depth.
Two additional bearings hold the guide-arm ass =mbly to the headwheel
panel. One is a bearing near the upper surface of :he panel and is pressed
into the panel and guide -arm shaft. A second bearing near the lower surface
of the panel is pressed to the headwheel panel and spring loaded to the
guide -arm shaft by means of a circular wedge on the shaft that is pressed
to the lower bearing by a spring and spring- retaining clamp on the shaft.
When the tape is moving, air pressure is applied to the guide air cylinder
(not part of the headwheel panel) which moves the guide -arm assembly
with an attached circular spring until the arm attached to the eccentric
shaft stops against the lead screw of the vacuum -glide positioner assembly
(not part of the headwheel panel) The position of the lead screw on the
.
EXERCISES
Q2 -1. Name the proper sequence of signal information as recorded from
"top" to "bottom" of a video tape.
Q2 -2. When the headwheel is mechanically aligned for minimum geometric
errors, is the headwheel itself physically moved?
Q2 -3. What is the head -to -tape velocity when the pole tips have a 2 -mil
projection?
Q2 -4. (A) What is the length of a tv line (1H, or 63.5 ,µs) recorded with a
head having a 2 -mil tip projection? (B) What is the length of H re-
corded with a head having a 1 -mil tip projection?
Q2 -5. If a recording is made with a pole -tip projection of 3 mils, can it be
played back on a head having only a 1 -mil tip without readjustment
to eliminate geometric errors?
Q2 -6. If a recording is made with a head having a 0.8 -mil tip projection,
will it meet SMPTE standards?
Q2 -7. What adjustment would be wrong in recording if there is a difference
in spacing for a tv line between the top and bottom of the video track?
Q2 -8. Convert 1.59 is to nanoseconds.
Q2 -9. If the raster width on a vtr picture monitor is 10 inches, what hori-
zontal displacement will occur between "bands" of picture with a play-
back error of 1 N.s?
Q2 -10. Why would the 4.2 -MHz portion of a multiburst signal break out first
at the vertical center of the image when the vacuum guide is adjusted
for lighter pressure (less tip penetration) ?
CHAPTER 3
stripped -off sync from the incoming video is utilized to obtain a 240 -
Hz signal (4 times a field pulse frequency of 60 Hz) , which is
amplified to obtain sufficient power to drive the headwheel. This
240 -Hz signal is also converted to a sine wave, which is recorded on
the control track of the tape.
6. Also during the record mode, a signal determined by the speed of the
headwheel is obtained and divided to 60 Hz to supply power to the
capstan motor. The speed of the capstan motor has absolute control
over the velocity of the tape, which is a nomin it 15 inches per second.
7. In the playback mode, the headwheel again rotates at 240 rps with
57
58 TELEVISION BROADCASTING
the initial reference being either the power -line frequency or local
sync. The 240 -Hz control -track signal is phase- compared to the play-
back speed of the headwheel, and any phase error slightly modifies
the 60 -Hz frequency applied to the capstan motor. Thus, the tape
velocity is continuously maintained so that the rotating heads sweep
directly across the video tracks originally laid down.
8. The audio signal is handled by conventional magnetic-recording
techniques. (The audio portion of the recording system is outlined
later in this chapter.)
O Supply Reel
0 Air Guides
Tonewheel and Head
Motor
Guide Shoe
s
Headwheel
( Guide Post
0 Simulplay Heads
Air Guide
1200 16 6
2400 32 8
4800 64 121/2
7200 96 14
mounted on the rear end of the capstan -motor shaft to smooth out
speed fluctuations.
12. The counting- capstan diameter and spur gears are designed to give
one count per 15 inches of tape. The counter indicates minutes and
seconds of tape, with a maximum time of 99 minutes and 59
seconds.
13. The lower (or right) stabilizing arm provides the same control
over the tape on starting the transport as does the upper arm (2) .
0I
FM Head
Video
In odulator Driver
r
Sync Strip
(60 Hz)
160 -Hz Ref
® Slip Ring -Brush Assembly
Headwheel -
®Servo Units 240 Hz ssii
0
Head Motor
x4
240 Hz
Capstan -
60 Hz 2$ Capstan
.4 Motor
60 Hz Drive Amp
Motor
(Derived From Head Speed)
2. The fm output from the modulator is amplified and split into four
separate channels, one for each of the rotating video heads. This is
necessary in the record mode because the video amplitude of each
head must be adjusted individually to obtain optimum results.
(NoTE: Whenever "video" is mentioned relative to the video heads,
it must be understood that the signal is in the form of a frequency-
modulated carrier.)
3. The individually adjusted fm outputs from the driver amplifier drive
all four video heads in parallel. No head-channel switching occurs
during the record mode of operation.
This completes the basic video function when recording. The remaining
functions serve in the timing of the recorded signals.
4. Sync pulses are stripped from the video signal being recorded, and
the field frequency (60 Hz) provides the -vitiating power source
for the rotating -head motor. This is termed the reference pulse.
5. The 60 -Hz pulses are multiplied to 240 Hz in frequency multiplier
62 TELEVISION BROADCASTING
A and fed into phase- comparison circuit B. The other side of this
comparator receives a 240 -Hz nominal frequency from the head -
shaft tachometer, a device for measuring shaft speed. The resultant
error signal is used to control the nominal frequency of 240 -Hz
oscillator C, which feeds the head -motor drive amplifier.
6. The head -motor drive amplifier is capable of supplying over 80
watts of 240 -Hz, three -phase power to the head motor. Nominal
power consumption is about 50 watts. The multiple -phase output is
obtained through the use of Scott-connected transformers.
7. The head motor is a three -phase 240 -Hz type.
8. The head -tachometer signal is used to generate the master -control
signal recorded on the tape and to develop the initial power refer-
ence for the capstan motor. This signal indicates the relative position
of the heads, controlled by the head motor, and the longitudinal
position of the tape, controlled by the capstan motor.
The tachometer (Fig. 3 -3) is simply a timing ring and a pickup head.
For example, the RCA "tonewheel" is a metal wheel with a notch that
breaks the magnetic path as it passes across the tonewheel pickup head,
inducing a pulse of current. The pulse is used to form a symmetrical square
wave of 240 Hz.
9. The control -track head records the signal (now a sine wave) which
is derived from the head tachometer. Since this same initial reference
from the head tachometer is also fed to the phase comparator (B of
item 5) , any tendency of the motor speed to "hunt" is compared to
the original signal timing to indicate the magnitude of error. The
corrective action of the servo controls the motor speed.
10. The capstan-motor drive amplifier receives 60 -Hz information de-
rived from the tachometer (through D and F of Fig. 3 -2) and pro-
vides amplification to the two -phase power necessary for the capstan
Timing Device on
Motor Rotating Shaft of
Headwheel Motor
Tachometer
0 Pickup Head
1/40
Second
NOTE: The foregoing discussion has omitted the audio function and the
circuitry associated with the vacuum -guide control. Audio is covered later
in this chapter, and the vacuum -guide circuitry is covered in Chapter 7.
64 TELEVISION BROADCASTING
Stripped Sync
Field -Pulse
Head -Motor To
From
Add Control Hea 'Motor
Incoming Video Gen Error
(or Local Sync
Circuits Drivers
Signal
for Local
Phase Gain
Recording)
60 pps Velocity Gain
240 Hz
Phase Velocity
Comp rator Detector
Lock Bus
Since the speed of the head motor (nominally 240 rps) determines
absolute values of deflection frequencies ( video scanning) , control -track
signal, and capstan speed, this reference (pulse from head tachometer) is
made as "tight" as possible relative to the recorded video signal.
Follow Fig. 3 -4. Pulses are formed from the stripped -off sync of the
incoming signal (being recorded) at the nominal field rate of 60 Hz.
These pulses feed one side of a phase comparator. The other side of the
phase comparator receives the nominal 240 Hz from the head tachometer.
This normally is a symmetrical waveform from a multivibrator controlled
by head speed.
A velocity detector also receives the nominal 240 Hz from the head
tachometer. The velocity detector compares this signal to a fixed reference
(not dependent on any external signals) which has a period equal to that
of the nominal 240 Hz.
Note that each leading edge of a head-speed pulse is equivalent to
one revolution of the head (360 °), or an average of 65.6 picture lines,
one-quarter of a field, or 4166 µs. But the square wave (as shown directly
BASIC REQUIREMENTS: WHAT THE SYSTEM MUST no 65
from the timing ring, Fig. 3 -3) that results gives a reference only for each
180° of rotation, or 33 lines.
The added outputs of the phase and velocity circuits of the servo feed a
head -motor control circuit which also receives the nominal 240 Hz from
the head tachometer. Head -motor control might be a voltage-controlled
oscillator or a headwheel "frequency" modulator.
The head -tachometer signal, which indicates the instantaneous speed and
phase of the head relative to the recorded signal, is fed to the control -track
recording head by way of a generator which mixes a 30 -Hz frame pulse.
This pulse is called the edit pulse, or the control -track frame pulse.
A 60 -Hz sine wave is derived from the head-tachc meter signal and used
to lock the capstan to the headwheel speed.
Now go to Fig. 3 -5. Waveforms 1 and 3 in Fig. 3 -5A show the lineup
of the leading edge of the headwheel -speed pulse with the fixed reference.
This is the function of the velocity loop in the headwheel servo. Waveform
2 is the resultant output of the head -speed multivibrator.
When the headwheel is slow (as in start -up) , the pulses from the tach-
ometer ( waveform 5, Fig. 3 -5B) will be farther aparr than the pulses of the
240-Hz reference (waveform 6) An error voltage which depends on the
.
difference between the two pulses is generated and Nerves to speed up the
headwheel. When the two pulse trains reach identi:al periods, zero error
voltage results, and the headwheel speed is maintained. This is done by a
pulse -sampling technique (Chapter 7) .
The velocity -lock condition (waveforms 7 and 8, Fig. 3 -5C) means that
an interval of about 33 picture lines is locked to a fixed reference within
the velocity loop. You can see that this is a "long-time- constant" afc loop.
Another way of looking at this is that the overall "horizontal rate" is an
integration of the horizontal rate of 33 picture line;.
When this lock is obtained, the phase loop is enabled to compare the
60 -Hz (field) reference from the signal being recorded to the horizontal
envelope of about 33 lines. This 60 -Hz sampling will then be on every
fourth pulse of the 240 -Hz signal. See waveforms 9 and 10 of Fig. 3 -5C.
This is the signal normally observed on the monitoring cro in the head -
wheel servo position.
360°
65.6H
I 0.25V I
Headwheel 1240 pps l
l 4166ps f
t i I
Velocity Reference I
1 Field I
16,667 ys
1440°
Headwheel at Start
Fixed ReferencelVelocityl
I--- 416616 -{ I
60 Hz
System Outputs
0 111
P ocessi g
Switcher FM Demod
Amp
240 Hz
(Gate)
Separated Sync
I
60 Hz Fr )m Places Switch
240 Hz
Head -Motor
Motor Local Sync Blanking Pulse Under
34, Drive Amp
Generator or Switcher Horiz
Power L ne Blanking
U
Tach
240 Hz From Head Speed 480 Hz
/Control -
Tread
\Head
-240 Hz
Phase Shift
(Tracking) 240 Hz
Osc
x4
=
240 x 2
480 Hz
Sauare Wave
m
Capstan
Servo 240 Hz
-Tape 240 Hz
Phase
O 16OHz Comp
Capstan 60 Hz
Cap tan
Motor o 0
Motor
24s Drive Amp
This ensures that the video heads "track," or pass directly over the recorded
video tracks on the tape. Failure of such control causer the relative positions
of the heads along the length of the tape to drift. Thus, a head could bridge
two adjacent recorded tracks rather than remaining over a single track.
A block diagram of the basic playback system is shown in Fig. 3 -6. In
the following explanation of the playback system, the numbers refer to
those used to identify the various blocks in Fig. 3 -6
1.The relatively low -level (about 2 to 5 millivolts) output from the
head slip -ring brush assembly is amplified in individual preamplifiers
before being fed to the switcher unit. For simplicity, the preamps
are not shown in Fig. 3 -6. The switcher combines the individual
head inputs into a sequential and continuous single output. A 240-
Hz signal from the head tachometer presets the gates. A 480 -Hz
signal, derived from the same source but with the leading edge
placed under horizontal blanking, is used to trigger the actual switch.
This technique is further explained in the discussion for item 14.
2. The fm demodulator extracts the useful video signal from the fre-
quency- modulated carrier.
68 TELEVISION BROADCASTING
5. The head motor is controlled by the same servo system that is used
in the record mode, except that the initial reference signal is the
local sync generator (or 60 -Hz power line) .
6. The head -motor drive amplifier serves the same function as for
recording.
7. The head motor functions the same as for recording.
8. The signal from the head tachometer is routed to units 1, 5, 10, and
13, as shown in Fig. 3 -6.
9. The control -track head reproduces the recorded control track.
10. The capstan servo compares the control -track signal (which is indi-
cating the relative position of the heads during recording) with the
head-motor tachometer signal (which indicates the playback speed
and phase) From this comparison, an error signal is developed
.
Time
0°
1
90°
2
180°
1 3
270°
I 4
360°
o-
1 po 240 Hz
Phase
O
1-
Gate Spl it
Gate 2 4 - in Time
29D° 11043 ys)
Phase
Split 90°
L
J
Switch
U On Off On Off o
4 270°
480 Hz
(B
Switch 1
1 Off On Off On
Switcher Output 1 I 2 I 3 I 4
-I I-1/960 Sec
-i-
Field Pulse Head Motor
Ph se HVelocity Control
Gn Detector
Comparator Circuits
60 Hz
240 Hz
Variable
De ay
240 Hz Fron Head Tachometer
60 Hz
1
TVA Contro -Track
Control -Track Tape Frame
Phase Puls Gen
Play (Head) Pulse
Comparator 240 Hz
-r1 Coinc
Vert
dente
Detector
Contro -Track
Phase
- +8
30 Hz
.F_
30 Hz
Reset Pulse
Sense
Field
Ref Frame 240 Hz Osc
Pulse Comp rator
Pulse NCO)
Gen Error
Volts
fi iv tH 1
Demodulated
Tape Sync
Input
Local
Composite
Sync
To
Capstan -Motor 4
Drive Amps
-
60 Hz s4
Counter
It is obvious that a 30 -ns jitter, when the picture is locked to the local
sync generator horizontally and vertically, is not observable. This amount
of jitter represents about one part in 2000 at the tv line rate of 15,750 Hz.
But what about the color time base? The color -sync burst contains a fre-
quency of approximately 3.58 MHz. This burst is the phase reference for
all the chroma information carried in the sidebands of the chroma sub -
carrier of the video signal. At 3.58 MHz, one cycle represents approximately
1
Then
280
0.8 ns per degree
360
Now, if you have a color -tape signal that exhibits as much as 30 -ns jitter,
there are 30/0.8 = 37.5° of burst jitter. Quite simply, this means the color
phase will jitter over 37° to 38° of the range, which is intolerable. So for
color, additional video time -base correction units are required. These units
are Color Time Error Correction (Colortec) for Ampex and Color Auto-
matic Time Correction (Color ATC) for RCA.
In practice, the monochrome error correction must be made first, then the
color error correction. Timing errors can now be reduced by this combina-
tion to about 2 ns, or approximately 2.5° of jitter at 3.58 MHz. Since the
jitter at the color -subcarrier time base must be about 5° to be noticed, the
result is quite satisfactory.
rapidly spoken words (or continuous music) so that proper picture con-
tinuity can be maintained.
Since the tape speed is 15 in /s, the spacing in time between the sound
and the corresponding picture frame is equal to (9.25/15), or 0.61 second.
It is obvious, therefore, that for most editing, the first frame pulse following
a spoken word may be selected for the splice (described in Chapter 10) .
Conventional audio magnetic-tape methods are used for the sound track.
A plot of the amount of remanence as a function of the strength of the
signal field is given in Fig. 3 -9. This type of transfer curve has no linear
portion that would allow an undistorted signal output. Therefore a high -
frequency bias (between 85 and 100 kHz) is added to the audio current,
as shown.
Ideal
(Linear)
Slope
Undistorted Output
AF
\ AF
Bias (Supersonic)
The audio component on the plus side of the zero magnetization axis
results in a distorted waveform in which loop A is larger than loop B. The
audio component on the minus side results in a distorted wave in which
loop C is smaller than loop D. The algebraic sum of A + C equals the
algebraic sum of B + D when the loops are combined on the axis of opera-
tion, resulting in an undistorted output.
The axis of easy magnetization of the iron -oxide particles on video tape
is aligned transversely across the direction of travel to favor video response.
This places a slight limitation on audio bandwidth and signal -to -noise
ratio, as compared to conventional 15 -in /s audio tape recording. It should
also be noted that the audio track on the video tape is only about 70 mils
(0.07 inch) wide as compared to the conventional quarter -inch audio tape.
74 TELEVISION BROADCASTING
Review Fig. 3 -1 for the audio head location. Fig. 3 -10 illustrates the
details of the post assemblies and the sequence of the upper and lower tape -
section heads. Note the dual function of each post. Note also that a cue -
track simulplay head is not provided. This position (bottom of third post)
is occupied by the control -track simulplay head. The control -track record
head is on the vacuum guide (Figs. 2 -24 and 3-1.)
Fig. 3 -11 shows the basic function of the rec /play changeover relay for
audio in any tape system. The relay contacts in this drawing are shown in
the play position. In this position, the contacts act as follows:
Tape Motion --a-
Audio Record/ Audio Simulplay
Audio Erase Play Head Head
1
F ) i
70 Mils
20 Mils
I r r I 1
1
Cue Erase Cue Record/ Control -Track
Play Head Simulplay Head
Audio
Simulplay
Head
Thb L0
6
7
Audio
Preamp
- Audio
Playback
Out
. o
Audio
Record /Play
Head
10
11
o
4 } 12 Bias
HF Bias
Audio
Equalizer Trap Control
Play
Record
F
Record Audio In
Fig. 3 -11. Audio record /play relay (playback position).
BASIC REQUIREMENTS: WHAT THE SYSTEM MUST DO 75
EXERCISES
The RCA TR -70 tape transport with head cover removed is illustrated
in Fig. 4 -1. The Ampex VR -2000 tape transport was shown in Fig. 1 -12.
Record
Control
Panel
Supply
'- " lr
r
Play
Control
Panel
Take -Up
Reel Reel
Air Selective Headwheel Audio/Cue Audio/Cue Audio and Capstan Pinch Counter Counter
Guide Erase Assembly Erase Red Play Control-Track Roller Roller
Head Head Head Simulplay Head
Fig. 4 -1. RCA TR -70 tape transpoirt.
back, the tape speed is controlled by the capstan; the supply and take -up
reels merely provide constant torque. In the wind mode, the tape is pulled
by the take -up reel, and the supply reel provides tension by pulling in the
opposite direction. The reel brakes are applied automatically whenever the
machine is placed in the stop mode. A foot -switch, located at the bottom
center of the front of the machine, permits the brakes to be released so that
the reels may be turned manually during tape threading.
The tension arms, located below each reel, take uá excess tape if speed
variations occur due to changes in operating modes,. Tape -break switches,
operated by either the take -up or supply tension arm, automatically stop
the machine if the tape breaks or if the supply reel runs out of tape. A
stationary, air-lubricated tape guide below the tension arm of the supply
reel positions the tape for its approach to the headwheel. The air guide
consists of a post that is perforated with tiny air holes and provided with
flanges at the edges. Air is forced through the holes to lift the tape slightly
off the surface. The flanges provide the guiding as Lion. (In some more
recent systems, the air holes have been eliminated, and rotating guides are
used instead.)
Tape from the supply reel proceeds under the tension arm, over the air
guide, past the tape lifter to a two- inch -wide selective erase head, and then
to the headwheel, which rotates at right angles to the tape motion. Below
78 TELEVISION BROADCASTING
the headwheel, a vacuum guide establishes suction which curves the tape
transversely to fit the wheel and holds the tape tightly so that the video
heads contact the tape. The guide position, which determines the head -to-
tape pressure, is determined by a lead screw driven by the vacuum -guide
positioner motor behind the tape -transport panel. The motor is controlled
by the guide servo system (described in Chapter 7.)
After leaving the headwheel, the tape passes under the control -track
head, which records or plays back a 240 -Hz sine wave mixed with narrow
30 -Hz frame pulses. The 240 -Hz signal is used during playback to establish
the timing of the capstan servo system so that the video heads scan along
the recorded tracks. The frame pulses permit locating the correct points at
which to cut the tape during tape splicing, and they serve other uses in the
servo systems.
Next, the tape passes under three supports in sequence; each support
contains two heads concerned with tracks at opposite edges of the tape. The
first support contains the audio /cue erase heads, which remove video signals
from the appropriate areas to provide space for recording on the audio and
cue tracks. The second support holds the cue and audio record /playback
heads. The third support holds the simultaneous playback heads which
permit monitoring the audio and control tracks during recording. Because
of the distance between the headwheel and the audio and cue heads, the
two sound tracks are recorded 9.25 inches ahead of the corresponding
video tracks. Located between the audio /cue record /playback head post and
the simultaneous -playback head post is the tape-edge guide post which
produces the required tape "wrap" around each head.
The tape then proceeds to the capstan and pinch roller. The pinch roller,
which is actuated by a rotary solenoid during recording and playback, pro-
vides friction which enables the capstan to pull the tape. The capstan motor
is driven synchronously by the capstan servo system so that the tape motion
is related to the headwheel motion. The tape speed is either 71/2 or 15
inches per second as determined by a speed selector switch. The capstan
motor drives a belt connected to the capstan flywheel and shaft.
After leaving the capstan, the tape passes under the counting roller and
the take -up tension arm to the take -up reel. The counting roller is geared
to a four-digit resettable counter which indicates elapsed time during
recording and playback and permits locating desired portions of the tape
during rewind. The counter indicates minutes and seconds and has a capa-
city of 99 minutes and 59 seconds, at a tape speed of 15 in /s.
It is probably apparent at this point in our studies that air systems play
a vital role in the tape- recording system. The description of a specific system
(RCA TR -70) will serve to give the student a practical view of the quad -
ruplex -tape air system.
THE TAPE TRANSPORT, AIR SYSTEMS, AND CONTROL CIRCUITRY 79
The high- and low- pressure air pumps and cooling blower are located
on the lower section of the machine, and are accessible from the front or
rear. Other parts are mounted on the front and rear of the tape -transport
panel and behind the play control panel.
Low -Pressure Air and Vacuum System
The low- pressure air and vacuum system (Fig. 4 -2) provides air for the
air -lubricated tape guide, vacuum for the tape lifter on the tape -transport
panel, and vacuum for the vacuum guide on the headwheel panel. Both the
air and vacuum functions are performed by a single pump of the oil -free,
rotary -vane type. The pump motor directly drives a slotted rotor which is
mounted eccentrically in a cylindrical air chamber. As the rotor rotates,
centrifugal force causes four carbon vanes on the rotor slots to press against
the cylinder wall. This action creates a vacuum at the pump inlet and drives
compressed air from the outlet. The pump is provided with a filter /muffler
and relief valve at the inlet (vacuum) , and a muffler, filter, and relief
valve at the outlet (air) .
Vacuum System -Air drawn in through the slots in the vacuum guide
on the headwheel panel flows through a short hose on the headwheel panel
to an outlet below the panel mounting plate. A long hose runs from this
outlet in the back of the machine, through a T connector on the rear shelf
frame, then to the vacuum filter/muffler behind the play control panel.
Vacuum passes through the filter/muffler and through a four -outlet con-
nector on its intake side. The vacuum gauge monitors the vacuum level at
this point, and a vacuum relief valve is connected to another outlet of the
connector to permit adjustment of the vacuum level. The final outlet of
the four-outlet connector is attached to a rubber hose that runs to the
lower section of the machine. Here, the vacuum l..ne is filtered prior to
being connected to the intake side of the low- pressure air motor or vacuum
pump. Thus, the vacuum guide obtains vacuum immediately when the low -
pressure air pump is energized, and it holds that vacuum in all operating
modes except when the motor is de- energized.
Exhaust Air System -Air supplied through the tir guide on the tape -
transport panel is fed from a hose attached to an outlet on the lower side
of the panel. The hose extends to a T connector on the rear shelf frame,
then to a low- pressure air filter behind the play control panel. After filter-
ing, the air is supplied through a four -outlet connector and to solenoid
80 TELEVISION BROADCASTING
r--
i
r---------
I
- - -1 0.
E
t,
r-----t
L__
T--I.
am* ION
Courtesy RCA
Fig. 4 -2. Low -pressure air and vacuum system.
THE TAPE TRANSPORT, AIR SYSTEMS, AND CONTROL CIRCUITRY 81
valve 1L1 (Figs. 4 -2 and 4 -3) The two remaining outlets of the four -
.
outlet connector are connected to a low- pressure ail gauge and an air relief
valve, respectively, so that air pressure may be monitored and adusted as
required.
Air solenoid 1L1 is energized in any running mode of the machine ( all
modes except stop, standby, or setup) and permits air to be supplied to the
air guide. When the machine is switched from a running mode to one
which stops the tape reels, a time -delay circuit in the control module de-
energizes the solenoid after a delay of approximately four seconds. This
Low -Pressure Low -Pressure High -Pressure Regulator Vacuum
Air Solenoid (1L11 L Air Gauge Adjustment Screw Gauge
Fig. 4 -3. Air system (front view) behind plag control panel.
delay permits the air to continue to flow through the air guide until the
tape reels have come to a full stop, thereby providing air lubrication at the
air guide while the tape is still moving.
An air muffler is attached to one side of the solenoid valve (1L1) to
reduce noise. The third outlet of the solenoid valve provides a connection
for the hose that runs from this outlet to the air exhaust filter on the air -
supply side of the low-pressure air or vacuum pump. With the motor of
the low- pressure air and vacuum pump energized in all running modes of
82 TELEVISION BROADCASTING
the machine, lubricating air is supplied to the air guide at all times that
the tape is in motion.
Tape Lifter -The tape lifter on the tape -transport panel is activated by
both air pressure and vacuum, which are supplied through solenoid valve
1L6 on the rear shelf frame. Each of the two T connectors mentioned pre-
viously in the descriptions of the vacuum and air flow have one outlet
connected to solenoid valve 1L6. Only during the wind mode of operation
is the solenoid valve set to permit exhaust air to flow through the solenoid
valve and into the tape lifter. This causes a sapphire rod to extend length-
wise from the tape lifter on which the tape rides and thereby prevent the
tape from passing across the selective erase head. When the machine is
switched to a mode other than wind, the vacuum line is connected through
the solenoid valve, and the air line is cut off from the tape lifter. The
vacuum presents an assist force to the tape lifter and causes the sapphire
rod to return immediately to its inserted position, thus causing tape erasure
to commence with minimum delay upon resumption of tape movement.
L_
L_
Courtesy RCA
Fig. 4-4. High -pressure air system.
84 TELEVISION BROADCASTING
high -pressure system is in operation; however, when the system stops, the
solenoid valve opens and permits any accumulated water to drain from the
air tank to a pan below the valve. Under normal conditions, the pan accum-
ulates a small quantity of water which is evaporated and does not necessi-
tate draining the pan.
Supply for Headwheel Air Bearing-High -pressure air is supplied from
the air tank through a rubber hose to a T connector behind the play control
panel. The T connector is attached to the high -pressure regulator, which
may be adjusted to the desired level. Two outlets are provided on the high -
pressure regulator. One outlet is connected to air switch 1S2. If the air -
bearing supply to the headwheel fails, this switch places the machine in
the stop mode to prevent the headwheel motor from being energized, and
the red AIR PRESSURE indicator above the tape -transport panel is illumi-
nated. The high -pressure gauge obtains its input on the opposite end of
the air switch.
The second outlet of the high -pressure regulator is connected to air
solenoid valve 1L4, which opens the air line to the output filter and head-
wheel air valve on the rear of the tape -transport panel. The output filter
removes water from the air system prior to supplying the air to the head-
wheel motor. Accumulated water is retained in the filter and the hose that
runs to drain solenoid valve 1L3 on the lower shelf of the machine. When
the machine operation stops, the solenoid is de- energized and permits the
trapped water to flow into a pan below the solenoid ( same pan is used for
solenoid 1L2) .
The compressor motor is energized by air switch 1S2 and air -bearing
relay 1K6 (not shown) , which applies operating power to the motor.
Vacuum -Guide Air- Cylinder Supply -Air from the tank also is fed to
solenoid valve 1L5, which is located on a rear-frame shelf behind the
machine. An adjustment valve attached to the solenoid permits the air
pressure to be adjusted to the correct level to set the operating speed of
the vacuum -guide air cylinder. This is accomplished with a needle-valve
control that is adjusted to barely allow air to pass. This needle valve should
never be fully closed, since air passage to the headwheel would then be
blocked. The output of solenoid valve 1L5 is supplied to the vacuum -
guide air cylinder, which positions the vacuum guide and causes tape
penetration by the headwheel according to the level determined by the lead
screw on the vacuum -guide positioner assembly. The vacuum -guide air
cylinder is located behind the tape -transport panel, near the rear of the
headwheel panel.
The motor that drives the low- pressure air and vacuum pump also drives
a blower which is mounted directly behind the air pump. This blower pro-
vides forced -air ventilation to the modules below the tape- transport panel
and to the voltage regulator on the lower shelf of the machine. Air enters
THE TAPE TRANSPORT, AIR SYSTEMS, AND CONTROL CIRCUITRY 85
the machine from the bottom shelf, through two flat rectangular filters
mounted behind the removable front cover of the lower shelf of the
machine. The filters are made with multiple layers of slitted and expanded
aluminum which is impregnated with adhesive to trap dust and dirt.
Another cooling blower is provided for the record -amplifier module and
associated filter. This blower is located adjacent to the headwheel blower,
behind the tape -transport panel and directly above the record electrical
filter.
Both blowers operate at all times that operating power is applied to the
machine. If the air supply fails for any reason, an associated switch auto-
matically places the machine in the stop mode. The passage of cooling air
through the blower air vent positions the fin attached to the switch arm
and holds the switch closed. If cooling air stops, the switch returns to its
normally open position and opens the operating power circuit.
The headwheel blower is used to cool the headwheel motor and to keep it
clean. The blower motor operates with ball bearings and is permanently
lubricated. The blower is a multistage type having an oil-wetted wire filter
in a single assembly.
Openings around the headwheel motor allow cooling air to be drawn in
and circulate around the motor. The air then is drawn through a flexible
pipe attached to the headwheel panel, and into the headwheel- blower
filter. Dust and lint are collected in the rotating filter before the air is dis-
charged around the blower motor.
master record mode. When the operator presses the MASTER button, a
momentary- contact switch triggers the master mode module, which latches
into the "on" condition and thereby energizes (grounds) a bus called the
master record bus.
The record bus, along with the output buses of the other mode modules,
is connected to the input of a diode matrix. This matrix may be represented
as two sets of buses crossing each other at right angles but not directly con-
nected to each other. In Fig. 4 -5, the input buses from the mode modules
are shown vertically, and the output buses are shown horizontally. Each
output bus goes to a load or group of loads performing related functions,
such as control of the headwheel motor, capstan motor, and capstan roller.
The loads consist generally of transistor switches, relays, or solenoids.
Where a solenoid is used, the output bus goes to a driver, or power -
amplifier, module which actuates the solenoid when the bus is grounded.
The driver modules are identical to the mode modules, but omission of a
feedback resistor in the external circuit causes them to act as amplifiers
instead of flip -flops.
The record bus is connected by an individual diode to each of the output
buses which perform a function required for entry into the master record
mode. Each diode is oriented to conduct when the record bus is grounded
and to be cut off when the bus is ungrounded. Similarly, the other input
buses are connected to their related output buses by diodes. Thus, when a
particular input bus is grounded, all related output buses become grounded
(through the diodes) , and all loads required for entry into the desired mode
are energized. This arrangement constitutes a simple, reliable control system
which uses no moving parts except the momentary- contact push buttons
which initiate the action.
When the machine is in the master record mode, it is undesirable to
permit direct entry into certain other modes (such as setup or standby)
without first going into the stop mode. To protect the machine from an
error by the operator, unwanted transitions are inhibited automatically by
transistor switches actuated either by the input buses to the diode matrix
or by one of the related output buses. These inhibit circuits apply a voltage
to the mode modules which control the unwanted modes. This voltage pre-
vents the machine from entering the undesired modes even if the operator
pushes the button requesting one of these modes.
When the recorder is in one mode ( such as master record) and the
operator presses the button of another mode which is not inhibited ( such
as the play mode) , the first mode module must be turned off to prevent the
machine from being in two modes simultaneously. To accomplish this, the
output of each mode module (except the audio and cue modules) is con-
nected to a bus, called the unlatch control bus, which goes to a transistor
switch or unlatch amplifier. The output of the unlatch amplifier goes to
another bus, called the unlatch bus, which is connected to the inputs of all
mode modules. When the unit is in master record, a voltage is applied
88 TELEVISION BROADCASTING
from the output of the master mode module to the unlatch control bus.
However, this voltage is insufficient to actuate the unlatch amplifier. If the
PLAY button is then pressed, the play mode module increases the voltage
on the unlatch control bus to a level sufficient to actuate the unlatch
amplifier. The unlatch amplifier then applies a voltage to the unlatch bus,
and this voltage turns off the master module. The play module remains on
because the trigger voltage overrides the unlatch voltage. Since only one
mode module is on, the voltage on the unlatch control bus decreases to the
cutoff level, and the unlatch amplifier goes off until the next mode is
selected.
The mode control systems for both Ampex and RCA also contain a tape -
motion detector and play control system that permits safe direct transition
from the wind mode to the play mode without having to actuate the STOP
button. Overstressing of the tape is automatically prevented by delaying
contact of the capstan roller with the tape until after the tape has come to
rest. This circuit also prevents transition from the wind mode to the setup
or record mode while the tape is in motion. These features are particularly
desirable in remote or automatic operation where it is not convenient to
ascertain when the tape has come to rest after winding.
EXERCISES
Q4 -1. Give the sequence of fixed heads (A) across the top of the tape and (B)
across the bottom of the tape.
Q4 -2. What determines head -to -tape pressure?
Q4 -3. What controls tape speed during record or playback operations?
Q4 -4. What is the function of the tension arms?
Q4 -5. What is the function of the tape lifter?
Q4 -6. What elements are supplied by the low- pressure air system?
Q4 -7. What elements are supplied by the high -pressure air system?
Q4 -8. In Fig. 4 -5, if the SETUP and MASTER RECORD push buttons were
depressed simultaneously, in what mode would the system be placed?
CHAPTER 5
This chapter will treat the entire video -fm system, bypassing circuitry
more appropriately termed "signal processing." Fig. 5 -1 shows the functions
to be studied in this chapter, while Fig. 6 -1 ( Chapter 6) illustrates the
signal -processing subsystem.
Playback Functions
l i
Video
Modulator
Demodulator
Clamps IFM
i
Modulator
AFC
r- FM Filter
t r--r ' i
w FM
Reference
I _ FM Equalizer
RCA
Only
Record Sy c
FM Switcher
Equalizer Demodulator
Record Headwheel
tilt
Playback
Amplifiers Panel Amplifiers
Post -Emphasis: This unit includes the necessary post- emphasis character-
istic for the demodulated video, and it also provides switching- transient
suppression and video line- output functions.
The video and fm system performs two major functions. These are:
(A) to record incoming video information on storage tape by means of
quadrature heads and (B) to retrieve video information from tape storage
(through the same heads) for playback.
A separate group of modules is used for each of the two functions.
Modules in the record function convert the incoming video to fm and
supply this fm signal through contacts on the record position of relays in
the fm preamplifier module to the quadrature heads on the headwheel.
During playback, fm is extracted from the tape by the heads, amplified
through the fm preamplifier, and supplied to the group of modules per-
forming the playback function for conversion back to video for further
use in the machine.
Special circuitry is included in the video and fm system to permit check-
out and adjustments in the system without the use of the quadrature
heads. This is the electronic -to- electronic (E -E) circuit, which connects
the record and playback functions back-to -back for setup adjustments.
NOTE: Fig. 5 -1 should be referred to often during the remainder of this
chapter to keep in mind the sequence of operations.
Sections 5 -2 and 5 -3 will consider the basic modulation-demodulation
process as typified by earlier (low -band) systems such as the Ampex
1000C and the RCA TRT -1A. A number of these systems are still in daily
use. Then, Section 5 -4 will consider the important refinements of the
combined modulation-demodulation process in most recent systems, typified
by the Ampex VR -2000 and AVR -1, and the RCA TR -70.
97 TELEVISION BROADCASTING
(C) Fm ri output.
Carrier'
Freq
- -
Sidebands of 5-MHz Carrier
,-t--
I
-L
Deviation Range
I I
I
I I RF Carrier
I I I
Video
13.6 j 3
I
2.5
I
1.5
I
0.5
I
l
r
I I
MHzI 4 j
I (
II I
I I
I
I I
White
SIB Tip -Sync Tip SIC
Mono
FM Output
Reference Tank Circuits
IApprox
for Frequency and
1 -8 MHz)
Deviation Indicator
oscillator signal, which is injected into the suppressor -grid circuit (Fig.
5 -5) The frequency of the master oscillator is 51 MHz, but it is not critical.
.
The operation of the master oscillator is the same as that of the fm oscil-
lator. The coupling between fm- oscillator stages (Fig. 5 -5) is through
common cathode impedance L8 and plate -to -grid coupling capacitor C17.
There are two heterodyne signals from the mixer; one is the sum of the
master -oscillator and fm- oscillator frequencies, and the other is the differ-
ence between the two oscillator frequencies. The output of the mixer is
applied to a two -stage amplifier (Fig. 5 -4) . Inductors in the control -grid
Video Input
Reactance Tube
22 pF
C17 p
1000
From Driver
Stage) n ?y
Nominal
Coupling Capacitor Common Center
Brought to Reference - Impedance - Frequency
Voltage During = 46MHz
Sync -Tip Time
Conducts - FM to
?RO
diode, and the amount of current is indicated on the meter. Maximum de-
flection of the meter is obtained when the frequency of the carrier output
equals that of the reference circuit selected. When S2 is in the Sync -Tip
position, positive sync -tip pulses force the gate tube into conduction,
thereby permitting the fm carrier frequency at the sync tips to be checked
with the other two reference circuits.
The Ampex 1000 -C modulator employs slightly different oscillator fre-
quencies, but the principle of operation (heterodyne to obtain carrier
frequency) is the same for systems following the original Ampex Model
1000 -A. The Ampex 1000 -C includes the Ampex standard monochrome
(somewhat higher pre- emphasis than RCA) and color networks, as well as
a considerably higher pre- emphasis termed Ampex Master Video Equaliza-
tion (AMVE) -any one is available by means of a push- button switch.
In addition, the equalizer components are plug -in units.
The Ampex 1000 -C modulator incorporates built -in crystal- controlled
test signal generators. Since carrier and deviation are set differently for
color than for monochrome, the modulator provides separate presetting
adjustments for the two. The recorder can be switched, either locally or
remotely, for either color or monochrome operation. The switchover
automatically actuates the correct preset carrier and deviation values in the
modulator.
The following crystal -controlled test frequencies may be selected by
push buttons on the Ampex modulator:
5.0 MHz-Monochrome blanking level
5.5 MHz-Color sync tip
6.5 MHz-Color peak white
6.8 MHz- Monochrome peak white
Two additional positions are also available on the Ampex. They allow the
user to plug in units for any special crystal-controlled frequencies.
The Ampex 1000 -C modulator employs a carrier frequency clamped at
the blanking level during the back-porch interval (Fig. 5 -6) The sync is
.
Z^
Sync
,¡Front Back
-H
Porch Porch
circuit of V12 passes through delay line DL1 before being fed to the
opposite control grid of V14. Delay line DL1 shifts the signal phase 180°
at 4.0 MHz, thus enabling the signals which appear at the control grids of
V14 to be in phase at that frequency.
Each section of phase splitter V14 supplies two signals 180° out of phase
to a pair of matched 6BN6 tubes (V15 and V16) comprising the dis-
criminator stage (Fig. 5 -8) The signal appearing in the plate circuit of
.
Input Frequency Slightly Above 4 MHz Input Frequency Slightly Below 8 MHz
Und GriSignal I
on Gri d I
H-0.125 vs-
Delayed and Inverted
Signal on Grid 1 JII
lil
II
Screen -Voltage
Waveform
U U (Inverse to Plate Waveform I
Tubes V15 and V16 may be considered as pulse -width gating tubes
because of their ability to produce pulses of varying width and constant
amplitude. The characteristics of these tubes are such that plate current will
flow only during the time a positive voltage appears at the control and
suppressor grids simultaneously. Therefore, the width of the negative pulses
appearing in the plate circuits of V15 and V16 is equal to the overlap
width of the positive pulses applied to their signal grids (Fig. 5 -9) . The
widest negative plate pulse occurs when the signals are in phase, and the
narrowest plate pulse occurs when the signals are oppositely phased. If the
suppressor grid is negative, the current is to the screen; if it is positive, the
current is to the plate. The duration of plate current in either tube is
directly proportional to the phase difference between the signals applied to
the control and suppressor grids. The net result is that the 6BN6 dis-
criminator amplitude-versus-frequency slopes are inherently linear.
Pulses appearing in the plate circuit of V15 and V16 do not occur
simultaneously, but are uniformly interspaced in midpositions because of
the phase -splitter action of V14. Thus, if the sets of pulses are well bal-
l0o TELEVISION BROADCASTING
anced, the combined gate output consists only of video information carried
by even harmonics of the fm carrier.
The characteristics of the 6BN6 tube are such that the sum of the plate
and accelerator currents tends to remain constant and independent of both
signal -grid potentials. Thus, a signal current in the plate circuit must be
accompanied by a nearly equal signal current of opposite polarity in the
accelerator circuit, thereby effectively providing a push -pull output from
each gate. The push -pull output signals are fed to cathode -coupled adder
stage V17B (Fig. 5 -10) . Note that the varying duty cycle from 4 MHz
to 8 MHz results in a dc level which represents the instantaneous fre-
quency -effectively converting frequency change to dc change.
Video -adder stage V 17B receives video information from the plate and
accelerator circuits of discriminators V15 and V16. This information is
added by V17B and appears at the plate. The video output is then fed into
a 10- section low -pass filter (DL2) which removes the carrier harmonics
and unbalanced fundamental carrier frequency above 5.2 MHz. When the
deviation of the fm carrier does not occur lower than 5.2 MHz, as in color
operation, this filter adequately removes small unbalanced carrier com-
ponents that would degrade the video signal. However, when deviation
occurs down to 4.3 MHz, with blanking at 5.0 MHz as in monochrome
operation, discriminator balance requires more careful adjustment to keep
blanking free of residual carrier not removed by DL2.
The video signal from filter DL2 is fed into video amplifier V18 (Fig.
5 -7) . In the plate circuit of this tube is a de- emphasis network which
matches the pre- emphasis circuit located on the modulator chassis. These
circuits serve to improve the signal -to -noise ratio.
Slightly Above 4MHz Slightly Below 8MHz
®
i
, l
O -
Sum of Both
O
U
Upper -Screen
Voltage Waveform J L
Lower -Screen
U Voltage Waveform
Fig. 5
®
Sum of Both
Screen Waveforms_ - .-
-10. Demodulator waveforms for both gate tubes together.
THE VIDEO AND FM SYSTEM 101
4.3 6.8
Monochrome
Deviation Range
E
Q
\ i
Cola Deviation Range
2 2.6
\ \\ \
\\
4
/ / //
5 2
5 5
6
6.5
MHz
Filter
Cu off
Fig. 5 -11. Video -sweep response of demodulator.
102 TELEVISION BROADCASTING
sent a 90° phase shift at the carrier frequency of 5 MHz (or 0.05 micro-
second) . The signal is split into two paths at the output of the last limiter
stage. One path proceeds directly to the grid of the adder tube, while the
other goes through the delay line before being fed to the adder stage. The
diagrams in Fig. 5 -13 illustrate what happens when the carrier frequency,
a lower frequency, and a higher frequency are passed through this circuit.
The resultant amplitude varies as a function of frequency; therefore, fm
demodulation is accomplished.
The am signal thus produced (which still contains an fm component) is
detected by a pair of diodes driven by a transformer, as shown in Fig. 5 -12.
The transformer acts as a phase splitter to provide the signal for full -wave
rectification, which is employed to achieve full -time signal recovery and
to cancel the unwanted second harmonics of the rf signal. After further
amplification, a low -pass filter eliminates all of the double frequencies
developed in the full -wave rectifier diodes and most of the carrier frequency
(filter cutoff is approximately 4.5 MHz) .
Filters Out
Fundamental of
Carrier
Detector
Video Low -Pass Balance
Amp
Out Filter Balances Out 2nd
Harmonic of Carrier
t
1
r1
Carrier Frequency Higher Than Carrier Lower Than Carrier
processes together. Study Fig. 5 -14 carefully. Fig. 5 -14A is a plot of signal
versus frequency of the video signal itself and the carrier and deviation
frequencies of low -band and high -band recorders.
The carrier frequency is deviated by the varying amplitudes of discrete
frequencies in the video signal. This produces upper and lower sidebands
with energies particularly strong at the second-order and third -order side-
bands (first and second upper and lower sidebands) .
3.58 4.2
.Color_
5.5 6 5
O O O O
I I I
I
I
I
-
I I I I
I, 1. 1 I
2 -1 0 1.8 2.4 3 6 6 0 90
Color Carri r Freq Carri r Freq
Foldover I
t Subcarrier (Low Band) (High Band)
Frequency (MHz) II
In low -band operation (for monochrome) even though the lower devia-
tion edge is adjacent to the highest- frequency video input component,
little effect is noticed in normal picture content. You will see the spurious
effect of this on the high- frequency wedges of a monoscope pattern, or on
the higher multiburst frequencies. It appears as a "beat" or moire inter-
ference on this type of test signal. But the monochrome program signal
has very low energy at the upper end of the passband, so very little sub-
jective interference can be noticed with this type of signal.
Now consider the low -band color operation. Here there is a discrete fre-
quency of high energy (the color subcarrier) beating with the instanta-
neous fm carrier. As deviation is increased, the sidebands increase rapidly
104 TELEVISION BROADCASTING
All circuits are now almost strictly solid state. Exceptions are in the first
stage of the video -head playback preamplifier (where nuvistors are used)
and in the video -head record drivers.
In solid -state modulators, the voltage -variable diode capacitor is used.
A review of this application in video modulators is presented in Section 5 -5.
Since this device has a nonlinear capacitance voltage relationship, the
result is about the same as that obtained with the average reactance circuit
(Fig. 5 -15).
F2 -F1= Fout
F2
Variable
Fin
F1
Fixed
Fout
The new dual heterodyne technique (Fig. 5 -16) overcomes this limita-
tion. One oscillator operates at about 100 MHz and the other operates at
about 108 MHz. Under modulation, they are deviated in opposite direc-
tions, with the result that cancellation of amplitude- frequency nonlinearity
takes place. The actual modulator frequencies depend on the particular
manufacturer and model number, but the principle is the same.
The difference frequency (modulated carrier) is obtained from the low -
pass filter (Fig. 5 -16). Video feed -through and even-order harmonics are
180° out of phase at the output, resulting in improved suppression. The
overall modulation transfer characteristic also results in about a 2-to-1
F3
Mod Carrier
increase in dynamic range. The main point to keep in mind is that the
variable- reactance diodes are arranged in complementary fashion so that
oscillator frequencies F1 and F2 change in opposite directions with respect
to the input signal.
P r
Junction III
Capacitance
Fig. 5 -18A illustrates the junction- capacitance curve typical of nearly all
signal diodes. Note that, in the reverse -voltage region of operation, the
capacitance change is quite small over a 100 -volt excursion. By comparison,
Fig. 5 -18B is an example (only) of the curve for one type of variable -
capacitance diode. There are many different types of variable-capacitance
diodes with different sensitivites in capacitance per volt of change. All of
these diodes typically have a nonlinear relationship between capacitance
and voltage change, as shown by this graph.
Some common circuits employ a voltage-variable diode capacitor, which
may be known as a 'variable- capacitance diode," "varicap," or "varactor."
This is a diode which is manufactured to maximize certain physical and
chemical features affecting junction capacitance. Fig. 5 -19 shows three
symbols for this type of diode that may be found on schematic diagrams.
THE VIDEO AND FM SYSTEM 107
2.0 100
1.9 80
a 60
U
2, 40
16 20
0 01 0 1 1 0 10 100 2 3 5 7 10
Like the zener diode, the variable -capacitance diode is always operated
in the reverse -biased condition. Capacitance decreases with increased re-
verse bias; that is, the capacitance varies inversely with the reverse bias
voltage. The capacitance- versus-voltage relationship is nonlinear. The diode
normally has a greater change in capacitance for a given change in applied
voltage than the effective capacitance change in a conventional reactance-
tube circuit.
When the total capacitance used to resonate a tuned circuit is split and
used to form a feedback voltage divider, a Colpitts oscillator results. Fig.
5 -20A illustrates a transistorized Colpitts oscillator, in which Cl and C2
Cathode
Anode
varies with the instantaneous video voltage. Since the diodes are effectively
connected to the tank circuit, the resonant frequency is varied in step with
the instantaneous video voltage variation. Let us see now in which direction
the frequency changes between sync tip and peak white. Since sync is the
most negative signal, the reverse bias is increased. Since capacitance varies
inversely with the reverse bias, capacitance decreases, increasing the output
frequency of the oscillator. The same reasoning tells us that as the signal
goes toward peak white, the frequency decreases. The result is a frequency -
modulated (but nonlinear) output.
Note that with the same type of transistor and the same polarity of video
input signal, the polarity of X1 and X2 could be reversed, with the reverse
bias now negative rather than positive. In this case, the oscillator would be
caused to decrease in frequency at sync tip and increase in frequency as the
signal swings toward peak white. If the same video signal is fed to both
inputs of the foregoing oscillator circuits, one oscillator increases in fre-
quency as the other decreases. This is a push -pull type of frequency modula-
tion which tends to cancel the nonlinearity of modulation resulting from
the nonlinear voltage-capacitance relationship of voltage -controlled capaci-
tance diodes. When this type of modulation is not used, a nonlinear
amplifier of characteristics inverse to those of the modulator restores
linearity.
Obviously, this type of application is not limited to the Colpitts oscillator.
Nor is it limited to the production of frequency modulation. For example,
this type of diode is used in voltage -controlled delay lines (in place of the
shunt capacitors) as applied to automatic time correction (ATC) in video
tape recorders (Chapter 6) .
A typical video -modulator circuit arrangement for a modern tape system
is shown in simplified block form in Fig. 5 -21. Pre -emphasized video is
supplied through the deviation control, which allows proper setting of the
THE VIDEO AND FM SYSTEM 109
tc
m LL
fi 1
o.
E
.-L
i ii 1
ALL
o.
E
E i
c)
v
U
Q
Fig. 5 -21. Typical video modulator arrangement for television tape systems.
110 TELEVISION BROADCASTING
residual high- frequency products that have not been cancelled in the
mixer.
Modulator AFC
All modern tape systems employ a means of automatic frequency control
(afc) for the video fm modulator. See Fig. 5 -22A. Pre -emphasized video
is fed to the modulator afc. The fm reference circuitry supplies crystal-
controlled reference frequencies. The output of the modulator is also fed
back to the afc circuitry.
Pre- Emphasized
To
Video
P Modulator Rec
Amps
Mod
(A) Application in system. Video 1.
AFC
FM Ref
(Rec Stds)
lus Sampling
Video L F VCO
Delay
Modulator AFC
1
11
+
_FL__ Freq Too High
DC Freq Correct
I(
JI°°
Clamp Pulse
+
J 1 us
Clamp Pulse
U --.1 1- -1.5us
In Fig. 5 -22B, note that each of the voltage -controlled oscillators (vco)
is returned to a clamping bridge. For the clamping -pulse duration (which
occurs at a 15.75 -kHz rate), each vco is returned to a reference voltage.
Since the effective capacitance of the varactor diodes depends on voltage,
this sets the frequency of the oscillators.
Observe at this time that there is capacitive coupling to each of the
clamped reference returns from point 1.
The afc essentially has three inputs: (1) a back-porch sample pulse
formed from separated sync from the incoming video, (2) the fin output
from the modulator, and (3) a crystal -controlled black-reference frequency
from the fm reference (or record standard) . The sampling pulse is around
3 -µs wide. The clamp pulses are formed from a trailing -edge delay circuit
so that they are delayed from the leading edge of sampling by about 1 µs
with a duration of about 1.5 µs.
White Reference
Low -Band Low -Band High -Band
Reference Mono MHz Color MHz MHz
Gray Reference
Blanking 5.00 5.79 7.90
The other reference frequencies shown in Fig. 5 -23 are available as test
frequencies in most recent systems. Also in these systems, you will find a
"white reference" pulse in the demodulation circuitry; this pulse permits
checking and setting the deviation for proper recording level.
THE VIDEO AND FM SYSTEM 113
eir
t
S
d
b
t
S
o
ce
a'
io
CO o C.
O O
o G ?Ñ
Ln
ó Ñ
the fm test panel. This selects a different resistor for the voltage divider
and provides a gray dc control for the modulator.
As mentioned earlier, the minus 20 volts supplied by one of the five
standards is also fed to the oscillator control diode matrix. This matrix
produces control levels which activate specific transistor switches so that
the precise reference frequencies required for the standard selected are
produced by the crystal -controlled oscillators. The correct switches are thus
preset in the gray, black, and white oscillators, so that merely the applica-
tion of power to one or more of the oscillator circuits is required to obtain
the correct gray, black, or white reference frequency from that oscillator.
The particular frequency and the associated transistor switch that is
selected, as well as the position of the TEST SELECT switch on the fm test
panel, are listed in Table 5 -1.
Standard
4 5
Test K3 1 2 3 525 625
Select K4 Function 525 625 525 High- High -
Position K5 Activated LB Mono LB Mono LB Color Band Band
Norm Black 5w Q6 Q7 Q8 Q9 Q10
Black Freq 5.0 MHz 5.54 MHz 5.785 MHz 7.9 MHz 7.8 MHz
or Off
White *Sw Q1 Q1 Q2 Q3 Q4
Freq White Freq 6.8 MHz 6.8 MHz 6.5 MHz 10 MHz 9.3 MHz
* *Input On - - - - - -
Gray Sw Q14 Q14 Q14 Q15 Q16
Noise On
Gray Freq 6 MHz 6 MHz 6 MHz 9 MHz 8.5 MHz
* *Res Off - - - - -
°Activated only in E -E mode or in play mode when white insert function is selected.
O*No reference frequency is generated in this mode.
In the absence of a negative control signal from the diode matrix, the
transistors are cut off by the positive base bias. This opens the series circuit
from the oscillator through the crystal, the switching transistor, and capaci-
tor C34 to ac ground.
To activate a crystal, the base of the associated transistor switch is driven
negative by an output from the diode matrix; the transistor switch saturates
and connects one side of the crystal to ground through capacitor C34. The
opposite side of the crystal is connected to a common bus by which all
crystals in the generator are coupled ( through capacitor C35) to the base
of oscillator transistor Q17.
A Colpitts configuration is used for the oscillator circuit, with the col-
lector of Q17 connected to ac ground. The selected crystal, which repre-
sents the resonant circuit in the oscillator, is effectively connected between
the base and collector of Q17 because of the saturated condition of its
grounded switching transistor. Oscillation is sustained by feedback caused
by the connection of the emitter to the capactive voltage divider consisting
of C37 and C38.
Diode clipper X20 -X21 limits the base voltage swing so that the output
level becomes independent of transistor characteristics. Bias resistors R81
and R82, with emitter resistor R83, establish the dc conditions in the
oscillator. The oscillator output is coupled through capacitor C40.
Power for the gray-frequency generator is supplied only when the noise
test mode has been selected by the TEST SELECT switch on the fm test
panel. (Refer to Table 5 -1 for the frequency obtained for each operating
standard. )
The black- frequency generator supplies the black- reference frequencies
used in the modulator afc module. It functions in a manner similar to that
of the gray generator just described, except that five oscillator output fre-
quencies can be obtained instead of three.
The white- frequency generator contains four switch-controlled crystals
that control the frequency of the white oscillator (Q5) .
Both the gray and black oscillators supply their outputs to transistor
switching circuits that permit only one oscillator output to be supplied at
any given time. Switch Q13 controls the gray -oscillator output, and switch
Q12 controls the black-oscillator output. Either switch is de- energized
according to the position of relay K3, which biases off either Q12 or Q13.
The unbiased output is then amplified through fm- driver subassembly Al
and supplied to the modulator afc module as the black or gray fm reference
frequency.
For the white insert and timing description to follow, see Fig. 5 -26.
For components discussed, see Fig. 5 -24. The white oscillator and crystal
switches remain inoperative until they are energized through relay Kl.
With reception of a white -insert signal from the fm filter module, K1
operates and connects operating power to the white oscillator and its asso-
ciated switches.
THE VIDEO AND FM SYSTEM 119
Sync
(For Time Ref) -4.8 ps
6H Pulse
P1 -29
Interval
Q48 -C
Back
Gate
-Porch
7
Pulses 1lJt -bus
P1 -27
r-
Delay MV --i F-12 Vs
Q46 -C
White -Insert
MV 37 ps-
Q51 -C
White -Ref
X1/X2
Output
FM
X3/X4
Output
FM + Ref
P1 -31
Demodulated
White Pulse From
Std Ref Fre4 /
Video
output amplifier (Q58, Q59) to supply an output of 0.5 volt peak -to -peak
to the fm filter module.
Negative -going reference vertical pulses are supplied from the linelock
module to the discharge switch (Q40) The latter produces ramp sweep
.
signals that are used during video -head resonance tests. The sweep level
is set by a front -panel control, which sets the current level from current
source Q41 to charge the ramp capacitor (C61) .
The ramp pulses formed across the capacitor are coupled to comple-
mentary emitter followers Q42 and Q43. The pulses are then supplied to
the modulator module where they substitute for video signals from the
video input module during head tests. The linear fm sweep provides re-
sponse data which are used during the head antiresonance adjustments.
Video pre- emphasis and de- emphasis standards are set by the SMPTE
de- emphasis curve of Fig. 5 -28A. The pre- emphasis curve is complementary
to the de- emphasis curve-simply turn the curve "upside down" and change
the dB values from minus to plus to obtain the pre- emphasis curve.
122 TELEVISION BROADCASTING
Playback -Amp
Head Resonance Curve
Response Curve
FM in From Mod
4 Rec
Contains Channels
\ Pre -Emphasized
Video
to Heads
\.......--
¡
Adj
"Crossover"
Freq r
Peak Adj High Med Low
ti
10 16
Frequency IMHz)
-i
From
Head Other
Channels
To
Demod
6 dB/ Individual
Antiresonant
Octave Head
Circuit
(C) Head -response compensation.
Slope Equalizer
Head
De- Emphasis
From Playback Amp Demodulator "Flat" Video
(Video)
via "Flat" FM
0 1.0
-2.0 0.8
_
8: -4,4 á 0.6
s HB
-8.0 0.4
LBM
-14.0 0.2
L BC
Standard RI : R2 TI us 12 us
Proper frequency and transient response of the system depend not only
on the magnitude of emphasis, but also on the shape of the curve.
Therefore, the slope of the curve is stated as a time constant.
The curves can be defined as the normalized impedance of the network
of Fig. 5 -28B. The SMPTE specifications can be simplified by putting
them in the more practical forms of Figs. 5 -28B and 5 -28C. The SMPTE
specifications are:
Low -Band Monochrome (LBM) :
T = 0.132 µs
X = 4.0
T=0.240,us
X=6.56
124 TELEVISION BROADCASTING
High -Band:
T=0.600µs
X = 1.5
where,
T is the time constant,
X is the ratio of R1 to R2 in Fig. 5 -28B.
you are willing to let the overall performance slip over a period of time.
(Chapter 12 treats maintenance procedures.)
Note that in the networks of Fig. 5 -29, capacitor C shunts R1 so that as
the frequency increases, the net amplitude to the common -base amplifier
increases. Exactly the same time constants are used as in the de-emphasis
network which complements the particular mode of operation.
But, as stated earlier, the actual values of components must be modified
for the type of circuit used. If you calculate the time constants in Fig. 5 -29,
you will see that the period is somewhat longer than that listed in Fig.
5 -28C. But note how closely the ratio of R1 to R2 holds in each case.
1110 1471 pF
(C) High band. Emitter
R2
Common -
Follower 407.8 Base Amp
271.9
Record Voltage
From
Record Amplifier
Low -
Capacitance Playback Amp
Coax r
Nuvistor Input
Rotary 20
Type 7895
Head Transformer
220
Damping
í2 Resistor Degeneration
mH
ill( Required) Resistor
(If Required)
Mark X Video Head -J
Assembly and Nuvistor
Preamplifier
L +84 V
-I1 V Play
+ 50 V Record o
LI -R 25 05, 06
Reproduced RF i
From Nuvistor Feedback Feedback
Preamplifier on o-e. Amplifier
Amp ifier
09
Amplifier TP1
Mark X Video Q10, 011 07,08
Head Assembly
(Factory -Set
C7 To Jl
Response) RF Output
sented by C1 causes a phase error in the low rf frequencies during the play
mode. However, this error is compensated in the individual playback
amplifiers.
The video playback amplifier and the video record amplifier are physi-
cally placed on the same etched board, but they are completely separated
electrically and are powered by separate power supplies. There are four
such etched boards (one for each channel); each is located in one of the
four head -channel units. Fig. 5 -30B identifies the circuit functions and
relationships of the playback preamplifier. Each of the playback amplifiers
receives the rf signal (after amplification in an individual one -stage
nuvistor preamplifier) reproduced by the video head with which it is
associated. Each of the four nuvistor preamplifiers is integral with the
Mark X head assembly.
The output of the associated nuvistor preamplifier enters the playback
amplifier at pin 2 by way of a low- capacitance coaxial cable of specific
length. The signal at pin 2 is applied to the emitter of Q11, which presents
a very low input impedance. Transistors Q10 and Q11 form a feedback
amplifier. The combination of the nuvistor and transistors forms a low -
noise cascode circuit, of which the nuvistor is the input section and feed-
back pair Q10 -Q11 is the output section.
The output of Q10 and Q11 is amplified by Q9 which produces a gain
of approximately 2. Components Ll and R25 reduce the current feedback
of Q9 at low frequencies, thus increasing the gain produced at these fre-
quencies. This action compensates for the low- frequency losses and phase
shift in the cathode circuit of the nuvistor stage. Between 1 and 15 MHz,
Ll and R25 provide phase compensation; at frequencies below 1 MHz,
they provide increasing gain as the frequency is lowered. Their principal
purpose is that of phase compensation to maintain phase linearity of the
reproduced video signal.
The output of amplifier Q9 is applied to the base of Q8, which with Q7
forms a feedback amplifier. This stage produces a gain of approximately 6.
Variable capacitor C7 is factory adjusted to eliminate gain variations of
fractions of a decibel at 15 MHz (arising from variations in transistor
parameters or tolerances of other components). The combination of the
nuvistor preamplifier, interconnecting cable, and playback amplifier is
factory set by means of C7 to be flat within less than 0.1 dB between 1 and
15 MHz. (This is typical of many later stages in the video signal path.)
The circuit of each video head includes its own characteristic inductance,
resistance, and stray capacitance plus the input capacitance of the associated
nuvistor preamplifier. This causes resonance within the bandpass of the
video head. The condition of resonance causes a change of group delay
which must be compensated. These effects of head resonance are removed
by the playback amplifier, the characteristics of which are made exactly
opposite to those of the head.
Fig. 5 -31 is a block diagram of the playback amplifier in the RCA TR-
70 equipment. To accomplish its functions, three major signal -flow paths
are used. Each of these is described separately below.
1. The primary path, through slope- amplifier transistor Q5, which pro-
vides head -resonance compensation, linear phase -amplitude response
correction, and agc.
2. The secondary path, through emitter follower Q7, which provides fm
level detection for display on the cro.
-§¡
E
ZgL
Courtesy RCA
Fig. 5-31. Block diagram of RCA TR-70 playback amplifier.
130 TELEVISION BROADCASTING
In the playback mode, the four video -head outputs must be sequentially
switched to minimize noise and cross talk and to prevent two heads from
132 TELEVISION BROADCASTING
combines the outputs of heads that are 180° apart on the perimeter of the
headwheel.
Playback 4-to-2 Switcher 2-to-1 Switcher
Preamp Amp
00 00
Combined
Channels
00
corn
Paired
Channels
Combined
OOOO
--r
RF
anca Out
Combined
Ampex Numbering
Channels
O RCA Numbering
00
Ofl
Playback
Preamp Amp
00 OO RF Switcher
Purpose
The fm switcher module combines the fm outputs of the four heads as
they sequentially scan the tape during playback. Precision switching in two
separate channels (video and sync) places switching transients where they
THE VIDEO AND FM SYSTEM 133
4 X 2 Coarse Switch
See Figs. 5 -33 and 5 -34. This circuit receives four fm inputs and one con-
trol input. The four fm inputs are supplied from the playback amplifier
modules. Each of the inputs supplied to the fm switcher module is a short
period of fm having a duration that corresponds to the time that the asso-
ciated head is scanning a recorded tape. The periods of fm output from
each channel occur sequentially and overlap in time, since one head begins
its passage over the tape before the previous one has completed its scan.
A 240-Hz tonewheel (tw) pulse is used for the control input signal (see
waveform A of Fig. 5 -34). This tw pulse is generated in the tonewheel
processor module (612) Its timing is based on a signal generated in a
.
pickup coil by a notch in the tonewheel as the tonewheel rotates with the
headwheel. The tw pulse thus has a definite relationship with the rotating
quadrature heads. The tw pulse occurs near the center of the active tape
scan by head 1.
Two gates are used in the 4 X 2 coarse switch to combine alternate input
lines. Channels 1 and 3 are combined in the diode gate using X19 through
X22. (See Fig. 5 -34, waveforms B, D, and I.) Channels 2 and 4 are com-
bined in the diode gate using X8 through X11 (waveforms C, E, and J) .
Switching between input channels occurs in each gate during the interval
when neither input channel is supplying a tape fm signal.
134 TELEVISION BROADCASTING
.74
>
ax
1L gé
QO
ao
VI
Wo
of RCA fm switcher.
136 TELEVISION BROADCASTING
2 has completed its traverse (2080 microseconds after the initial switch) ,
the gate switches from channel 2 to channel 4. Thus, disturbances in the
desired fm signal are avoided by gating the two input signals while both
are inactive.
A similar sequence of operation occurs in the gate for channels 1 and 3.
However, in this gate the tw pulse cannot be used directly to trigger the
FM
TW
240 Hz
Channel
FM
Input
Input
Channel 2
1
'
! siiiiii k
pr viriivii
Y
..Mni
rATA"
Input %G
Channel 3 &NM NM
FM Input
Channel 4 WAYA ' WA WAN
/i.
214 Control
.
2080µs
Q43 Out
i i
Delay Pulse
1040 µs
Q45 Out
fM
1/3 Control
Q46
FM 1/3
Out
OI
-
Mn.
-
\
Gate Out
v, v
%'/iii,y °
M3 & M9 In
FM 2/4
Gate Out (
---
M2 &M9In
FM Video
%/l
Ch Sw Out
M2/M3 Out
M2 Anode
M3 Cathode®L
480 Hz
-
OK /X///i %//3iii7 %/.iiR'//
Fig. 5 -34. Timing diagram for 4 X 2 coarse switch and video channel.
THE VIDEO AND FM SYSTEM 137
control multivibrator (Q46 and Q47) This would cause switching to occur
.
during the active head traverse because the tw pulse is generated near the
center of the active scan for head 1. To overcome this problem, the tw
pulse is delayed for a time interval corresponding to 90° (1040 micro-
seconds) of rotation of the headwheel by delay -generator transistor Q45
(waveform G in Fig. 5 -34) The trailing edge of the delayed output pulse
.
from Q45 then triggers the delayed 240 -Hz multivibrator (Q46 and Q47,
waveform H) and causes the gate for channels 1 and 3 to switch between
the intervals when heads 1 and 3 are supplying fm.
Output from the gate for channels 1 and 3 ( waveform I) is fed through
buffer transistors Q21 and Q20 to one input of the video -channel fm switch
and one input of the sync -channel fm switch, respectively. Similarly, the
output from the gate for channels 2 and 4 (waveform J) is fed through
buffer transistors Q9 and Q8 to the second inputs of the video -channel and
sync -channel fm switches.
The delay generator (Q45) provides an auxiliary output for the fm
switcher module through pin 10 to the Cavec module ( 626) when the
latter accessory is used with the machine.
Video -Channel FM Switch
Two diode quads constitute an electronic switch controlled by precisely
timed signals developed in the timing -generator circuits of the fm switcher
module. One input to the electronic switch is the combined fm from chan-
nels 1 and 3 at quad diodes M3 (waveform I of Fig. 5 -34) The second
.
The outputs of the two quads are connected in parallel to drive subse-
quent stages. The control points of the quads are connected in complemen-
tary fashion to the push -pull control signals so that the quads select the
inputs alternately as a single -pole, double -throw switch. Switching occurs
at each polarity reversal of the control signal, so the switch operates at a
960 -Hz (4tw) rate when supplied with a 480 -Hz control signal (wave-
form L) .
The 960 -Hz switching rate causes the quad diodes to select the channel
1 or 3 input while head 1 is traversing the tape. The quad diodes then
switch to the channel 2 or 4 input, just as head 1 is leaving the tape and
head 2 is beginning its scan. As head 2 completes its scan, the quad diodes
switch again to the channel 1 or 3 input to select the output of head 3 as it
starts across the tape. When head 3 is nearing the end of its active scan,
the quad diodes switch away and select the channel 2 or 4 input once again
as head 4 begins its traverse. Finally, as head 4 nears completion of its scan,
the quad diodes return to the channel 1 or 3 input to begin the sequence
again with head 1.
The output from the quad diodes is a continuous fm signal (waveform
K in Fig. 5 -34) resulting from the sequential selection of playback chan-
138 TELEVISION BROADCASTING
nels coordinated with the passage of the heads across the tape. Because the
headwheel rotates 240 times per second, complete cycles in the selection of
all four heads must occur at a 240 -Hz rate, and selection of individual heads
occurs at a 960 -Hz rate.
Video -channel fm in its entirety (waveform K) is fed to a diode switch
(X4 and X5) , which is controlled by the record bus. This switch is open
for operation in the record mode to prevent high -level recording fm, which
may be picked up in the sensitive fm preamplifier module, from proceeding
through the fm switcher module. Thus the switch eliminates possible cross
talk in later stages.
Fm through the switch is supplied to a feedback line driver (Q6 and
Q7) that restores the signal level lost in the switching circuits and the
output termination. A high -pass filter is used between the line driver and
the output terminal (pin 17) to allow usable fm to continue to the fm
equalizer module (523) , but to reject low- frequency offsets caused by
minor dc unbalance in the 2 X 4 gates and M2 -M3 switches.
Sync -Channel FM Switch and Equalizer
Refer to Figs. 5 -33 and 5 -34. Diode quads M9 and M10 in the sync
channel function in the same manner as M2 and M3 in the video channel,
except that timing of the switching process is advanced in the sync chan-
nel. However, unlike the arrangement in the video channel, the sync-
channel fm does not pass through a separate amplitude -equalizer module
prior to demodulation. Instead, playback losses are compensated for by
simplified equalizers with fixed adjustments in the fm switcher module.
Two separate equalizers in the module are optimized individually, one for
high band and the other for low band. Relay Kl selects the proper equalizer.
It is energized by standards bus 4 or 5 for 525- or 625 -line high -band
standards, and it remains de- energized for all low -band standards.
Either equalizer operates in conjunction with common -base amplifier
Q40 to boost high- frequency fm signal components relative to the low -
frequency components. Sync- channel video is used only to provide a source
of tape sync undisturbed by switching pulses; therefore, adjustment of the
equalizers is not so critical as in the video channel, and fixed controls can
be used.
Equalized fm is fed through buffer emitter follower Q39 to the feedback
line driver (Q37 and Q38) . The fm is then filtered through a high -pass
filter as in the video channel.
Timing Generator
The complexity of the timing -generator signal flow requires the descrip-
tion to be in three parts, each having a seperate block diagram and timing
diagram.
Video Channel, Normal Operation-For this description, refer to Fig.
5 -33, the simplified diagram in Fig. 5 -35, and the waveforms of Fig. 5 -36.
THE VIDEO AND FM SYSTEM 139
Sync-
Inverter
Channel G30
Switch
4TW - !
960-Hz
Bistable MV
C141, 042
EF
Q16
12yu
Pulse Gen
Q15
Video - InputGate
Channel (AND)
Switch SI Q14
l
480-Hz To Video -Channel
TN/ Bistable MV Switch Driver
Q!, Q2
pulse is inverted by transistor Q30 and triggers the 960 -Hz bistable multi -
vibrator (Q41 and Q42) An additional trigger pulse for the multivibrator
.
2 ps Division
Sync Channel
Switch Pulse
4TW
1-
r
Adjustable With 41W De ay
960 -HV MV
Q42 Output
Pulse Gen 12 us r
Q15 Output
Video Channel
Switch Pulse O
I put Gate
n
Q14 Output I
FM
N.
M2 Input
M3
FM
Input & N. \, 1
Head Overlap'
%. 3
FM
M21M3 Output
. \
FM Interval Corresponding
Q2 Output
O to Horizontal Sync
M2 Cathode
M3 Anode O ¡(
M2 Anode
M3 Cathode O I
1040 ys/Division
FM 214
Gate Output if0 /AM /AM 'MW OEM %/í.L/
M2 Input
FM 113
Gate Output
H NIKV
/
M3 Input
FM
M2 /M3 Output
I
4 WASFA MM VA 1
Q2 Output
480 Hz
M2 Cathode (
M3 Anode
480 Hz
M2 Anode
M3 Cathode
O
240 HzTW
tive -going transition timed during the horizontal -sync interval at the
precise instant that video- channel head switching is to take place. This
signal is connected to the second input of the input gate when switch S1
is in the Normal position.
The input gate produces a positive output when both inputs are negative.
This condition exists during the time interval between the leading edge of
the 15,750 -Hz video- channel switch pulse and the end of the 12- micro-
second 960 -Hz pulse (see waveform F, Fig. 5 -36) . In summary, the char-
acteristics of this video-channel trigger pulse are determined as follows:
1. Leading edge: Video- channel switching pulse
2. Trailing edge: Sync -channel switch pulse plus a 12- microsecond
interval generated in transistor Q15
3. Repetition rate: 4tw (960 Hz)
A positive -going output from transistor Q14 triggers bistable multivi-
brator Q1 -Q2, which operates as a two -to -one frequency divider. The
square-wave output from Q1 -Q2 is at a 480 -Hz rate (waveform J, Fig.
5 -36) and is clipped by shunt limiter M4 (Fig. 5 -33). The signal is then
supplied to the video- channel switch driver, where push -pull control signals
(waveforms K and L, Fig. 5 -36) are developed to drive the video- channel
precision switch (M2 and M3 in Fig. 5 -33) The latter supplies a continu-
.
4TWo -
-33, 5 -37, and 5 -38. Playing back or recording a tape using the rf -copy
960 -Hz
Bistable MV
041, Q42
.a EF
016
12 µs
Pulse Gen
015
n pu Gate
Inverter I
(AND)
Q30 o 014
SI
Sync -
Channel o
Switch
To
Horiz -Gate 480-Hz Video -Channel
I nverter Pulse Gen
Delay Bi stable MV Switch
019 017
018 01,02 Driver
TW o
Fig. 5 -37. Simplified diagram of video -channel timing generator, rf -copy mode.
142 TELEVISION BROADCASTING
Video Channel
Switch Pulse O
Sync Channel
Switch Pulse
Inverter
Q19 Output
Adj by
Sw Delay-1
Horiz
r
Gate Delay OD l'
QI8 Output
Pulse Gen
Q17 Output
E
-- - - - - -- fi -1/ _---- - -i_ -J
Pulse Gen ( 7)
Output O 12 ps
Input
014 Output
Gate
G
s.. -a. -1
- - - -I` -_ - -_.
FM
M2 Input
Head Overlap*
O
¡
FM
M3 Input
Exact Position of Sw tch Depends
on Switch Delay Setting
FM
M21M3 Output 3
J
FM Interval Corresponding Position of Switch During Sync
to Front Porch Interval in Normal Operation
A new, adjustable delay pulse is derived from the sync -channel switch
pulse (waveform B of Fig. 5 -38) , which occurs almost six microseconds
in advance of the video-channel switch pulse ( waveform A) . The sync -
channel switch pulse is inverted by transistor Q19. The resulting output
pulse drives horizontal-gate delay generator Q18. Here, a new delay pulse
is generated; its leading edge is coincident with the input pulse, but its
trailing -edge transition time ( waveform D, Fig. 5 -38) is determined by
the position of sw DELAY potentiometer R85.
The variable -width pulse from transistor Q18 drives pulse generator Q17,
where a new constant -width pulse (waveform E) is formed. The leading
edge of this pulse is coincident with the trailing edge of the input pulse.
Thus, the position of the pulse is determined by the setting of the Sw
DELAY control.
Pulse generator Q17 supplies the variable-delay pulse at a 15,750 -Hz
rate. The pulse is combined with the 960 -Hz, 12- microsecond pulse (wave-
form F) from pulse generator Q15 in input gate Q14. (The pulse from
Q15 is generated exactly as in normal operation.) The input gate produces
a positive pulse (waveform G) each time the inputs from Q15 and Q17
are both negative.
Sync- Inverter
Channel ----jl 030
Switch
960-Hz
4TW Bi stable MV -a EF
Q33
Q41,042
1W
480-Hz
Bistable MV
Q31, Q32
-- ToSync-Chan nel
Switch Driver
The resulting pulse from the input gate has a positive -going leading
edge that triggers the 480 -Hz bistable multivibrator (described previously) .
However, the multivibrator output transitions occur during the front -porch
interval, and quad diodes M2 and M3 are switched accordingly. Thus, the
output of the quad diodes (waveform J) is a continuous fm signal repre-
senting sequential switching between heads, with the switching point
occurring in the front -porch interval as adjusted with the SW DELAY
control.
Sync Channel-Refer to Figs. 5 -33, 5 -39, and 5 -40 for this description.
The 4tw input pulse (waveform B, Fig. 5 -40) causes the 960 -Hz bistable
multivibrator to change state as a head is just beginning to traverse the
tape. After the multivibrator has changed state, the next sync -channel
144 TELEVISION BROADCASTING
switch pulse to occur passes through inverter Q30 and returns the multi-
vibrator to its original state, where it remains until it is reset by the next
4tw positive trigger. This results in a 960 -Hz negative pulse from the
multivibrator (waveform C, Fig. 5 -40). The leading edge of this pulse is
determined by the positive -going 4tw pulse transition, and the trailing
edge is timed by the sync -channel switching pulse.
Sync-Chan (7C\
Pulseo
4TW
960-Hz MV
Q42 Output
480-Hz MV
Q32 Output
M9 Cathode
E
M10 Anode
M9 Anode
t
MIO Cathode I
Input
OM9 FM : Head Overlap
nput
M10 IFM
H
FM
M9/M10
Output
OI ZARIBBarA
Picture Interval
\L\
Horiz- Blanking Interval
blanking.
Head -Switch Pulse Generator
Gate Q10 (Fig. 5 -33) is enabled by the record bus whenever the
machine is in any mode other than record. At those times, the gate passes
480 -Hz pulses from the 480 -Hz multivibrator (Q1, Q2). Each transition
of the 480 -Hz signal triggers the monostable multivibrator (Q11, Q12),
and the multivibrator generates seven -microsecond pulses at a 960 -Hz
rate. This is now the head- switching pulse output, the leading edge of
which coincides with the video-channel switch point. During input or
noise test conditions, the post- emphasis module utilizes the head -switch
pulse to clamp out switching disturbances in the demodulated video signal.
the video input module is now removed, and the video is restored to its
original form. A separate section in the module is used to process the sync -
channel signal similarly, so that both channels maintain separate character-
istics as their signals are supplied to the video processing system ( next
chapter) and the monitors.
Sync-channel demodulation (Fig. 5 -1) is performed with the sync-
channel demodulator module (used by RCA) , which is identical in con-
struction to the video -channel demodulator module. Using a separate
channel for sync demodulation provides a desirable enhancement of picture
quality and sync stability. This is accomplished by having sync-channel
switching occur during fm overlay corresponding to the picture interval.
Thus, timing errors are not introduced, and switching disturbances do not
occur during the demodulated sync interval. Conversely, the video-channel
switching is accomplished during sync -interval fm overlap, leaving the
active picture interval undisturbed. The disturbance to the sync pulse
caused by switching is overcome by completely replacing the sync pulse
with a clean pulse derived form sync demodulation in the processing
amplifier (Chapter 6) .
NOTE: In the Ampex system, the processing amplifier derives and pro-
vides switching pulses which cause video head switching to take place
on the front porch of sync, thus avoiding any ambiguity in the timing
edge of horizontal sync.
See Fig. 5 -41. In the video -channel demodulator, the signal is split in
phase and processed in a push -pull manner through the module. Limiter
stages clip the signal to remove variations in level, and push -pull fm is
supplied to the discriminator stage.
Conversion of frequency variations to amplitude variations is initiated
in the discriminator circuit where constant -width pulses are generated at
a repetition rate corresponding to the instantaneous carrier frequency. The
effective video level from the discriminator is very low compared to the
carrier level. However, the filter following the discriminator partially re-
moves those carrier components that may cause cross talk, and it supplies
the push-pull video to an amplifier stage to provide the module output.
Push-pull output from the demodulator module is supplied to the video
filter module, where selection is made for the proper Bode -type filter
according to the operating standard of the machine. The Bode filters present
a characteristic that passes video and attenuates fm. The constant-width
pulses from the demodulator discriminator are integrated, forming video
that passes relatively undisturbed through the passband region; however,
in the stopband region, the phase is caused to lag by 90° in one half of the
filter and lead by 90° in the opposite half of the filter. This lag and lead
condition occurs at higher frequencies, which is where fm cancellation is
required.
THE VIDEO AND FM SYSTEM 147
+12 V
Cascode
Diff Amp
+12 V
Phase
Splitter
Diode Detector
1I
Cascode
5 Similar
Differential Diff Amp
Stages
Amp
J
To Filtering
8 Output
Limiter 1
& Emitter
Follower
ll
Diff Amp
Fig. 5 -42. Simplified schematic of delay -line drivers and diode detector.
150 TELEVISION BROADCASTING
EXERCISES
Q5 -1. Under what circumstances could electronic "quadrature correctors" be
eliminated in both the record and playback functions?
Q5 -2. What is the fm bandwidth of (A) older tape systems and (B) later -
model combination high -band, low -band systems?
Q5 -3. Give the corresponding frequency for each of the following:
(A) Low -band sync -tip level (monochrome)
(B) Low -band blanking level (monochrome)
(C) Low -band peak white (monochrome)
(D) Low -band color sync -tip level
(E) Low -band color blanking level
(F) Low -band color peak white
(G) High -band sync -tip level
(H) High -band blanking level
(I) High -band peak white
Q5 -4. Define "dropouts."
Q5 -5. Approximately how much rf voltage exists at the headwheel brushes
in (A) the record mode and (B) the playback mode?
Q5 -6. What is the primary signal that supplies switching information to the
electronic channel switcher?
Q5 -7. Where does the input signal to the demodulator come from in (A) the
record mode and (B) the playback mode?
Q5 -8. Why is fm amplitude limiting used in the demodulator?
Q5 -9. Why must timing information from separated sync be fed to the
electronic switcher?
Q5 -10. Is head -resonance compensation used for fm equalization?
CHAPTER 6
In the internal mode, ATC corrects geometric distortion but does not
synchronize vertical- or horizontal -sync pulses from the tape recorder with
the corresponding sync pulses from the local sync generator. To accomplish
this type of correction, the reference sync is timed by deriving an average
from a long- time -constant afc loop locked to tape sync. A dc error signal
is obtained by making a phase comparison between a sample pulse
derived from the trailing edge of tape sync and a trapezoid waveform
generated from reference sync. The trapezoid timing is steady enough to
eliminate geometric distortion but changes sufficiently with the average
timing of tape sync to prevent slow drifts from causing the error signal to
exceed the correction range of the delay line.
Table 6 -1. Servo Terminology
Mode
No. Ampex RCA Servo Mode
In the external mode, essentially the same circuits are used, but the afc
loop is locked to local sync instead of tape sync. Since the local sync signal
is fixed in timing, the error signal derived by comparing it with tape sync
contains information to correct both jitter and geometric distortion. As
a result, the ATC greatly reduces residual Pixlock (automatic) jitter and
provides an extremely stable output.
THE VIDEO PROCESSING SYSTEM 155
In either of the two modes, the error signal is amplified and applied to
the delay line. As the demodulated composite video signal passes through
the delay line, the length of the line is modulated by the error signal every
63.5 microseconds, or once per tv line.
Full correction range is maintained automatically by a tha ( tape horizon-
tal alignment) feedback circuit. This circuit adjusts the relative phasing be-
tween the ATC sample pulse and the ATC trapezoid so that the average
error signal is in the center of the delay -line range. In the internal mode,
the tha signal controls a variable delay in the internal reference afc loop.
In the external mode, the tha signal causes the Pixlock servo to advance
the phasing of tape sync with respect to local sync by controlling the head-
wheel motor speed.
Geometric distortion includes various segmentation errors which result
in picture distortions such as jogs, scallops, and quadrature -error indenta-
tions (Chapter 2) Segmentation errors are sufficiently minimized for
.
throughout the line, increases at a constant rate and is the same for cor-
responding points on successive lines.
Errors resulting from scalloping cannot be removed completely by an
ATC system, because the timing error at the right-hand side of the picture
is not the same for successive lines. The difference in timing errors for
successive lines will be seen in the ATC output as a small skew -like effect
in vertical lines at the right -hand side of the picture only. It is, therefore,
156 TELEVISION BROADCASTING
the errors of Fig. 6 -2A. (This is the condition of Fig. 2 -14, Chapter 2.)
Waveform 1 is the sync time recorded, and waveforms 2 and 3 show the
playback sync timing for the first and last lines of a complete band of 16
lines. Waveform 4 shows that this results in a series of linearly progress-
ing errors across each band of 16 lines, or the duration of one head pass.
Now go to Fig. 6 -2B. Since errors must be sampled (measured) before
a correction is made, a fixed delay of 2 to 3 µs is employed, as shown, prior
to application of the corrector voltage to the voltage -controlled video delay
line.
Reference sync changes with the mode of operation. For nonsynchronous
operation, this reference is separated horizontal sync from the video signal
after demodulation. For synchronous operation of the servos, the reference
becomes local horizontal sync from the station sync generator. The termi-
nology here is "internal" for nonsynchronous modes and "external" for
synchronous modes.
In the internal mode, tape sync supplies both the sampling pulse and the
trapezoid. The trapezoid is obtained by applying tape sync to a long -time-
constant afc loop. This mode provides a relatively large range of control,
since the trapezoid changes sufficiently with average timing of tape sync
to prevent drifts from exceeding the characteristic impedance (Z0) of the
variable delay line.
When the voltage -controlled variable delay line is modulated away from
its center of operation, Z. becomes increasingly mismatched (mistermi-
nated) When a given range of correction (delay) is exceeded, reflections
.
from mistermination occur and upset the servos. The maximum of all errors
in the time base should not exceed 1 µs.
THE VIDEO PROCESSING SYSTEM 157
Uncorrected Video
Controls
Tape Sync Snc
Phase in Sep
Synchronous
Mode Fixed Ref Volts
Head Error
Motor Det
i 1
H ad Trap THA
Servo Gen Diff Amp
In Out lot
Dellay
Delay /
L
Sample
i
+ Ext
Ref
Error Volts
Since a large timing shift may occur when the recorder switches from
one head to another during playback, it is essential that timing errors be
measured after switching so that the proper amount of corrective delay is
provided for the next line. In older systems, switching took place during the
horizontal -sync pulse, and the trailing edge of tape sync was used for
deriving the timing -error signal. In the latest systems, switching takes
place at a predetermined time ahead of the leading edge of sync, and the
leading edge of tape sync is used for deriving the timing-error signal.
Operation of the electronically variable delay line depends on the fact
that the capacitance of a reverse -biased silicon diode (varicap diode)
changes with applied bias. A variable delay line could be constructed
simply by connecting a number of inductors in series, connecting one
diode from each junction of the coils to a common bus, and applying a
variable dc voltage between one end of the series inductors and the com-
mon bus. However, this design would not be satisfactory because the
varying bias used to change the line length would appear in the output
signal. To overcome this, a second set of diodes is used, and each diode of
this set is connected from one of the junction points to a second bus (Fig.
6 -3) A fixed negative dc voltage is applied to the center (inductive) por-
.
tion of the line. A variable positive voltage is applied to the common bus
of one set of diodes, and a variable negative voltage is applied to the com-
mon bus of the other set. The bus supplying the positive voltage is called
the positive error bus (PEB) , and the bus supplying the negative voltage
is the negative error bus (NEB) The dc levels of the PEB and NEB
.
signals are fixed to equal amounts above and below the voltage applied to
the center, and ac components are equal but oppositely phased. The video
input is applied between one end of the coil series and ground, and the
delayed video output appears between the other end of the series and
ground.
THE VIDEO PROCESSING SYSTEM I?`)
Ma
It
.:.
r:.
"--{ 1
I
KM"
L:.
r:
t:
mElt.
160 TELEVISION BROADCASTING
Microseconds
o 1 2 3 4 5 6 8
I I I I
Re(
Horiz Sync
(System
01 I
Sync Input)
Demodulated
Horiz Sync O r
ATC
Sample Pulse
(From Tape
Horiz Sync)
ATC
Trapezoid
(From Local
Horiz Sync)
Since the delay line has a nonlinear relationship between the voltage ap-
plied and the time delay, a linear signal cannot be used to drive the delay
line directly. To overcome this difficulty, the error signal is applied to an
amplifier whose output varies nonlinearly with the input signal (Fig.
6 -5A). The amplifier is carefully designed to match the delay -line character-
R2
3. 3
3. 2
3. 1
2. 9 I
I
I
I
2. 8 I t
I I
2. 7
I ¡ Minimum Delay
2. 6
2.0 S
05 10 15 2.0 25 30 35 40
PE BI NE B Volts
PEB & NEB 0;25 1.1
Volts
I I 1 I I I I I
NLA Out 0.5 1.0 2.2 3.0 4.0 5.0 6.0 7.0
Volts
I I I
istic so that the overall relationship is linear. The output of the nonlinear
amplifier is applied to a phase splitter which provides the PEB and NEB
signals, and through a fixed delay line to another phase splitter which pro-
duces the DPEB and DNEB signals. Each of the four error signals is
applied through an individual complementary-symmetry emitter -follower
stage to the variable delay line. Circuits of this type are used because they
162 TELEVISION BROADCASTING
provide low output source impedance on both positive -going and negative-
going signals to the error buses.
To recapitulate, the picture -channel video from the demodulator is fed
to the monochrome ATC video module. The video is amplified, clamped,
and applied to an electronically variable delay line. The video signal is
either delayed or advanced from the nominal error delay of 3.05 micro-
seconds ± 0.05 microsecond (Fig. 6 -5B) The delay line is controlled by
.
the positive error bus and negative error bus (PEB and NEB) signals,
which are developed from a phase comparison of the tape sync and the
reference sync in the ATC error detector module. The output of the MATC
video module is fed to the Color ATC video module where further error
correction takes place. Essentially, this is the most important function of
the MATC system. All of the other circuits in the system only contribute
to the detection of an error between tape sync and reference sync and the
controlling of these auxiliary circuits.
When the tape recorder is operating in the tonewheel servo mode, the
tha error signal is fed to the ATC trapezoid circuit in the ATC error
detector module, where it shifts the phase of the ATC trapezoid waveform
with respect to the ATC sample pulse. The tha loop is stabilized when the
PEB voltage is equal to the tha reference voltage of +1.1 volts. Under
these conditions, the dc average of the error signal keeps the electronically
variable delay line in the center of its range. It should be noted that this
condition is necessary to obtain maximum correction range (Fig. 6 -5) .
ber of samples per line due to the presence of approximately eight cycles
of subcarrier in the burst, and because of the possible use of a narrower
bandwidth.
A second factor o'hich prohibits the use of horizontal -sync phase com-
parison in color operation is that there is no specified phase relationship
between the edge of sync and the color subcarrier. In addition, it is possible
that time modulatior may be introduced between horizontal sync and the
color subcarrier; therefore, horizontal sync cannot be used in controlling
the absolute phase of a subcarrier. It is essential that the final color timing
correction be accomplished by comparing the phase of burst with a fixed
reference.
The MATC open -loop system is capable of maintaining an error- reduc-
tion factor of at least 25 to 1, which is sufficient for monochrome reproduc-
tion but may leave enough residual error to disrupt color performance if
large errors are to be corrected. Because of the cascading effect of the
CATC process, residual error theoretically can be reduced by another fac-
tor of 25 to 1, and any error remaining is then negligible, even for color
operation.
Color ATC systems employ the same open -loop control described above
for monochrome signals; however, in color systems, the error detector uses
burst instead of horizontal sync. Also, CATC systems require a delay range
of only 360° at the subcarrier frequency (0.28 microsecond at a subcarrier
frequency of 3.58 Mhz); therefore, the delay line is shorter than that
utilized in the MATC systems and need not be split to contain the delay -
change transient within the horizontal -sync interval. The CATC nonlinear
amplifier, fixed delay line, and driver circuits are identical to those of
MATC systems.
In CATC systems, it is necessary that the phase -error detector have the
equivalent of a linear delay range of 360° at the subcarrier frequency.
Design limitations prevent the detector from functioning effectively over
an entire 360° delay range; therefore, the detector is operated at half-
frequency, and the required 360° delay range appears to the detector as a
180° range. Multivibrators divide both the burst and reference subcarrier
signals; the divided burst is used to form narrow sampling pulses, and the
reference signals form a sawtooth waveform. A special inhibit circuit in
the burst path insures that the narrow sampling pulses always occur on the
center half of the sawtooth slope.
The control signal for CATC variable delay lines is formed in the color
phase- detector circuit by the measurement of burst phase error. The
function of the control signal is to correct phase errors in composite color
signals and stabilize signal timing in order to provide excellent color repro-
duction. However, color signals must still be passed through processing -
amplifier circuitry to clean up blanking and sync and insert a new color
burst that is free of noise. This requires special circuits which do not
distort the color components of a signal.
164 TELEVISION BROADCASTING
the error buses. However, since tape burst information from which the
sample pulses are derived is obtained before the video signal has been
delayed by the color fixed delay line, transients applied to the variable
delay line advance 2.5 microseconds with respect to the video signal applied
to the line. Transients occur during the horizontal sync pulse interval and
do not affect burst or video information in a composite signal. The output
of the variable delay line is a composite video signal which is stabilized in
time with respect to the reference subcarrier. This signal is fed to the color
processing system.
Circuits of the color processor separate low- frequency and high- frequency
components of a delayed video signal so that they can be processed inde-
pendently before recombination. The separated high- frequency chroma
signal is fed to special processing circuits in the color processor module.
Special processing circuits for the high- frequency signal are required be-
cause the monochrome processing-amplifier circuits clamp out the burst
signal, and the black clipper removes all signal components below black
level. The low- frequency signal is fed to the MATC reference module
(Fig. 6 -6) Circuits in the MATC reference module separate sync from
.
the low -pass signal, and the separated sync is used to develop regenerated
sync in the sync logic module (Fig. 6 -1) Regenerated sync is fed to the
.
video output module where it is combined with the low -pass and high -pass
portions of the video signal (processed in the video control and color
processor modules, respectively) to form the video output signal.
THE VIDEO PROCESSING SYSTEM 165
As shown in Fig. 6 -1, the reference 3.58 -MHz subcarrier signal is fed to
the color phase module, which contains two electronically variable delay
lines. The first variable delay line is controlled by the SYSTEM PHASE
control and can be varied over a nominal range of 0.3 microsecond ( i.e.,
a delay equivalent to a subcarrier phase shift in excess of 360° ). Following
this delay, the reference subcarrier signal is fed simultaneously to the
second variable delay line in the color phase module and to the color error
detector module. The second variable delay line is controlled by the BURST
PHASE control, and the delay range available is also nominally 0.3 micro-
second. The reference subcarrier delayed by the second variable delay line
is designated the burst subcarrier and is fed to the color processor module.
The reference subcarrier fed directly to the color error detector module is
designated the system subcarrier.
Tape burst, separated from the video signal in the burst processor
module, is also fed to the color error detector module. In the color error
detector module, the system subcarrier signal is used in generating a refer-
ence sawtooth waveform; the tape burst signal is used in developing a
sample pulse. The sample pulse and sawtooth waveform are fed to a
phase-comparator circuit where the sample pulse falls higher or lower on
the sawtooth waveform slope, depending on the relative phase between
the reference -subcarrier and tape-burst signals. The output from the phase-
comparator circuit is an error signal, the magnitude of which reflects the
difference in phase between the system (reference) subcarrier and tape-
burst signals. Logic circuits insure that sampling occurs only over the
linear portion of the sawtooth waveform slope; therefore, the gain of the
error detector is linear. If the sample pulse falls near either extreme of the
range on the sawtooth waveform slope, it is possible to obtain a cracking
effect ( i.e., successive pulses sampling at alternate extremes of the slope,
which causes conflicting error information). To prevent this effect during
normal tape playback in the CATC mode, a portion of the phase -error
signal is used in controlling the thaf (tape horizontal alignment, fine)
signal in the CATC system; the thaf signal in turn adjusts the video delay
slightly so that the sample falls near the center of the sawtooth slope.
The burst logic circuit, which controls the sampling process, is in turn
controlled by the narrow burst -gating pulse developed in the burst
processor module. The width of this pulse is variable; however, the width
is nominally set at that value which allows approximately five samplings
during each line. The output error signal fed to the Color ATC module is
a fluctuating series of dc potentials. To eliminate the possibility of an
erratic output signal such as could occur if tape burst were suddenly lost,
a clamping circuit provides a steady dc output signal which corresponds
to zero error if the tape burst signal is absent. The clamping circuit is
controlled by the burst -lock sense potential which in turn is determined
by a logic circuit in the color processor module. The phase -error signal fed
to the CATC module is compared with a fixed reference in a differential
THE VIDEO PROCESSING SYSTEM 167
amplifier circuit, and the resulting output is developed into the PEB and
NEB potentials which control the variable delay line.
The PEB and NEB potentials which control the electronically variable
video delay line are developed from the phase -error signal of the color
error detector. The phase -error signal is an indication of the relationship
between the reference -subcarrier and tape -burst signals and varies linearly
with the phase difference between these signals. However, the character-
istics of the video delay line are such that the delay varies nonlinearly in
response to a linearly varying control potential impressed across it. It is
desired that the delay vary linearly; therefore, it is necessary to convert the
linearly varying phase -error signal to a nonlinear potential having gain
characteristics which match the delay characteristics of the video delay line.
The nonlinear conversion is accomplished by a differential amplifier circuit
that incorporates a gamma network containing resistors with values calcu-
lated to produce a nonlinear output which compensates for the nonlinearity
of the video delay line.
The nonlinear error signal is split in phase, and signals of opposite
polarity drive complementary- symmetry emitter followers which produce
PEB and NEB potentials. It is important that the PEB and NEB potentials
be equal in amplitude and opposite in polarity with respect to the voltage
to which the video signal in the delay line is referred, over the entire
range of the phase error signal. Therefore, potentiometers are provided as a
means of adjusting the PEB and NEB potentials to insure that a change in
one is accompanied by an equal and opposite change in the other.
Potentiometers are also provided as a means of adjusting the transient
behavior of the PEB and NEB potentials so that the rise times are as nearly
equal and opposite as possible in response to a step of error such as occurs
at switching when a vacuum -guide error is present.
Fig. 6 -6 shows the tha and thaf closed feedback loops to the headwheel
servo. The block diagram shows the functions of the headwheel servo,
MATC system, and CATC system that operate in conjunction with the tha
and thaf closed loops. In general, the principle function of the tha loop is
to adjust automatically the phase of the headwheel servo to maintain the
average sampling of the MATC in the center of the error detector slope.
This point, in turn, is set by internal adjustments to correspond to the PEB
potential for the center delay of the monochrome variable delay line.
The principal function of the thaf loop in the CATC system, during
playback in the normal CATC mode, is to adjust automatically the phase of
the MATC output signal to maintain the average sampling of the CATC
in the center of the error detector slope. The center of the error detector
slope in the thaf loop is also set up by internal adjustment to correspond
to the PEB voltage representing the center delay of the color variable delay
line. The actual operation is accomplished in several steps. First, the thaf
error signal is generated by comparing the PEB potential to a reference
voltage. The resulting error signal is applied, through a low -pass filter, to
168 TELEVISION BROADCASTING
lines. When the output is not in coincidence with reference vertical, the
vertical coincidence circuit will drop out and thereby throw the servo out
of the Pixlock mode. To prevent the machine from falling out of the Pix-
lock servo mode during color tape playback, tape vertical is timed to the
output signal by the fixed delay inserted into the tape vertical path.
In the nonphased color (npc) mode of tape playback, requirements are
somewhat different than in the normal CATC mode. The principal function
of the npc mode is to allow playback of noninterlaced color (nic) tapes.
Noninterlaced color tapes are produced by dubbing a playback from a
heterodyne color system. Inherent in the heterodyne color system is an
output signal in which the phases of the monochrome and chroma portions
of the signal are displaced by an amount equal to the jitter of the tonewheel
servo in the machine that made the original recording, with the result that
they no longer have a fixed relationship. The Linelock and the MATC
circuits, which are attempting to correct the video- signal time base by
comparing tape sync to reference sync, and the CATC system, which is
attempting to correct the video -signal time base by comparing tape burst
to reference subcarrier, are apparently presenting conflicting information.
However, the presentation of conflicting information is avoided by causing
the CATC system to assume control.
As the relative phase between the tape burst and reference subcarrier
attempts to change, the thaf differential amplifier senses this condition and
generates a signal which modulates a delay in the path of the reference sync
fed to the Linelock and MATC circuits. The headwheel servo and MATC
systems follow this change in phase of reference sync in such a manner that
the phase of the tape burst does not change with respect to the reference
subcarrier. The bandwidth of the thaf loop during color tape playback in
the npc mode must be greater than that of the tonewheel servo used in
recording the original tape. The net result is that the output signal from
the machine varies in phase in such a manner that the chroma signal bears
a fixed phase relationship to reference subcarrier. That is, the output signal
is noninterlaced color, but it is as satisfactory as the original noninterlaced
color playback.
To accomplish modulation of the reference horizontal delay with as great
a range as possible, it is necessary to provide a nominal delay of approxi-
mately one-half line. The nominal delay can then be modulated ± 20 micro-
seconds, if necessary, to follow the jitter of a very loose servo. A similar
fixed delay of approximately one -half line is inserted into the tape hori-
zontal path, so that the undelayed tape signal retains approximately the
same phase relationship to the undelayed reference sync during playback
in the npc mode that it does during playback in the normal CATC mode.
This allows the headwheel servo to operate in the Pixlock mode during
color tape playback in the npc mode, since vertical alignment is maintained.
Even though the tape signal is varying in time in response to modulation
of the reference horizontal delay, if its nominal position is centered with
170 TELEVISION BROADCASTING
respect to reference vertical, the excursions do not exceed the range of the
lock -sense circuit. In tape recorders equipped for the Linelock-only (LLO )
servo mode, it is advantageous to operate in that mode during npc playback
when vertical coincidence is not required. Since exact horizontal phasing is
not predictable in playback of a noninterlaced color tape (nor is it precisely
adjustable in playback of an interlaced color tape in the npc mode) , there
is no possibility of mixing, fading, or incorporating special effects, etc.
The prime purpose of the normal CATC mode is to provide these possi-
bilities during playback in the Pixlock servo mode.
Delayed reference horizontal, with the same modulation, must also be
fed to the MATC system. If the MATC system were operating from a fixed
horizontal reference, it would attempt to correct the timing modulation
which has been inserted into the tape signal by the headwheel servo. How-
ever, the delayed reference horizontal is delayed by nominally one-half line,
whereas the tape signal is nominally aligned to undelayed reference hori-
zontal. It is necessary to insert another half-line delay in the reference-
horizontal path to the MATC system to insure that the trapezoid waveform
is properly timed so that the sample pulse falls on its slope. This nominal
half -line delay is automatically adjusted to be precisely the correct amount
to cause average sampling at the center of the slope by the tha error signal.
It should be noted that the tha loop is relatively limited in bandwidth so
that the ATC is still performing a normal function of removing timing
errors such as quadrature error, jogs, etc. It should also be noted that the
thaf loop, during npc playback, is also bandwidth- limited so that it does not
respond to the normal line -by -line correction of phasing errors by the
CATC system.
There is one further refinement of MATC operation required for color
tape playback in the npc mode. This refinement consists of bypassing the
afc loop to form the ATC trapezoid directly from the delayed reference
horizontal. The afc loop performs no useful function in the external mode
of MATC operation; it is actually detrimental during color tape playback
in the npc mode, because it provides undesired bandwidth limiting and an
additional phase shift which makes the high -gain thaf loop very difficult
to stabilize. During color tape playback in the normal CATC mode, how-
ever, the afc loop provides a required timing advance which is not required
during npc operation because a delay of one full tv line is provided.
The remaining circuitry of Fig. 6-1 associated with the ATC system prior
to video processing -the tape sync processor and internal reference -will
be examined in the following two sections.
'0
58
Q Ó
35
w Ó
and (6) a 6H pulse. The tape sync processor module also derives horizontal
and vertical sync signals during back -to -back (E -E) operation.
As illustrated in Fig. 6 -7, tape horizontal sync is stripped from the
incoming video signal and is gated, at the horizontal rate of the system,
to two outputs by the leading -edge gate. One output is sent to the internal
reference module for use in the afc loop timing; the other is sent to the
ATC error detector module for use in MATC timing. Tape vertical sync
is stripped from the incoming video signal and is sent at the vertical (field)
rate of the system to the Linelock module, where it is used in the tva and
tm vertical circuits (Chapter 7) . The tape frame pulse is generated by
gating the tape horizontal and tape vertical sync pulses and sending the
resultant, at the frame rate (one-half the field rate) to (1) the capstan
phase module, where it is used in the capstan servo operation, and (2) the
splice logic module, where it is used for Switchlock detection (Chapter 10).
The back-porch clamp pulse is derived from the trailing edge of the
incoming composite-video sync signal and is sent, at the horizontal rate
of the system, to the MATC video module, where it is used to establish
the average picture level (APL) , and to the composite trailing- edge-gate
pulse generator to reset the flip -flop. The 6H pulse output, at the vertical
rate of the system, is fed to the fm reference module, where it is used to
control the white insert function (Chapter 5).
The tape sync processing circuitry for both Ampex and RCA systems is
described further in Chapter 7 under Servo Functions.
Er :is
a =
-r-
ES
-- ^
o
Y-T-L
8
--1,
fi c;cGD
ÉtiÓ ö5,
Psz
g-- >s E
J®
Ó rarfg
t
i4
Sync -
Processor -
TP2
AFC Tape Horiz Horiz Rate
Osc Output
Q10 Emitter
4 ps
r
---4.6 ps
h
AFC Trap
Delayed
Tape Horiz
-I II- 4-5 ps
A4-11
IJ
-I I-- 3.2 ps
Q15 Collector
1-2ps
Pl -29 Video -Channel Switch
1
P1-30 8 ps Trailing -Edge Gate
SwitchaOle)
9 ps-I
PI-32 Leading -Edge Gate
I- 3.2 ps
PI-27 AFC Hori./
j-4ps
PI-28
Head Switch Suppress
The output pulses are delayed from the input horizontal-sync pulse by
various amounts determined by their functions. Refer to the timing diagram
in Fig. 6 -9.
Internal reference functions in both Ampex and RCA systems are dis-
cussed further in Chapter 7 under Servo Systems.
r NO n
ó m Ñ
E 8"-
E,"
oapiA passaooud ya
o^'
o
O N
áá
-6
>
>
ow(S ua6aa au(c ua6aN
a
E
U
3,"
01
2 CO
y
>
E
U
U
HZ
aA >
ó
2 E
U
OS Ind l-e1S
Pulse Processing
The three pulse- forming modules that carry out the sync timing and
regeneration functions (Fig. 6 -11) are the horizontal afc module (534) ,
the sync logic module (533), and the vertical advance module (532).
These three modules are closely interlinked functionally and physically to
produce composite blanking, gated horizontal sync, and regenerated (out-
put) sync. It is important to note that the blanking pulses needed for both
monochrome and color tape-recording systems must be regenerated from
the tape sync pulse and hence must be advanced in time relative to the
leading edge of sync.
Start Pulse
Clamped composite video from the video processor module is fed to the
horizontal afc module. Vertical sync is separated from the composite video
and integrated. The integrator circuit has a long time constant and only
functions during the vertical -sync interval. The output of the integrator
produces a start pulse which is coincident with the leading edge of the
second vertical -sync pulse. The start pulse is fed to the vertical advance
module, where it starts the field -rate counter. This circuit counts pulses from
the master oscillator in the horizontal afc module. These pulses occur at a
rate of twice the horizontal line frequency (2 X 15,750 = 31,500) and
therefore with a separation of one-half the tv line time (63.5/2 = 31.5
microseconds) .
CO
E
U
%.
CL
Fñ
U
CL
á
trol, since for a small change in time a large change in voltage is produced.
The trapezoidal waveform is produced by gating the master oscillator and
horizontal -blanking pulses and feeding this pulse to a trapezoid generator.
The sample pulse is produced by the separated composite tape sync.
The sampling medium is a diode quad which acts like a normally open
switch that is closed by the sample pulse. During the pulse time, the closed
switch applies a voltage determined by the height of the trapezoid slope at
that instant to a long- time -constant RC circuit. This RC circuit smooths
out rapid timing variations so that the oscillator is not constantly shocked
by quick changes. During times when the tape sync and afc sync pulses are
not locked, the time constant is much faster, and the master oscillator
responds to changes more rapidly. This permits a faster lock -up of the afc
loop.
Blanking
Composite blanking is made up of horizontal blanking from the hori-
zontal afc module and vertical blanking from the sync logic module.
Composite blanking is applied to the video processor module so that old
sync may be removed from the composite video signal and new sync added
later in the video output module. The horizontal-blanking pulse is used
internally in the horizontal afc module to gate the delayed afc sync pulses.
Sync Selection
Three sources of composite sync from the sync logic module are available
to be added to the video output. They are (1) regenerated sync, which is
new sync made from the leading edges of the afc sync pulses from the
horizontal afc module, (2) processed sync, which is made from the leading
edges of tape sync pulses, and (3) house sync, obtained directly from the
local station sync generator.
180 TELEVISION BROADCASTING
Clamp Pulse
Stripped tape sync, from the video output module, and horizontal blank-
ing, from the horizontal afc module, are fed to the clamp -pulse generator.
The output of the generator is fed to the video processor module where it
clamps the video circuits.
Video Processing
The two video processing modules that form the video adder and sync
adder functions are the video processor module (432) and the video output
module (433) shown in Fig. 6 -12. These two modules are closely located
physically so that the video signals are not affected by external noise.
Processed Video
Composite video from the CATC video module is applied to the video
processing module, where it is amplified. The gain of the amplifier is
controlled either on the front panel of the video processor module or on a
remote panel.
The amplified composite video is split into a monochrome channel and
a chroma channel. The clamp pulse generated in the sync logic module is
applied to the back porch of the horizontal interval in the monochrome
channel and clamps it to -10 volts. At this point, composite blanking is
added, and the old blanking and sync are removed. A new pedestal level
is added; it also is variable from a remote panel, if selected.
Meanwhile, the chroma portion of the video signal is applied to a clamp
circuit which removes the old burst signal only when the tape recorder is in
the play mode. Chroma gain is adjustable only at the local location on the
front panel of the video output module.
The monochrome and chroma channels, along with the regenerated
color -burst signal, are combined in the video adder. The output of the video
adder (at this point video without composite sync) is applied to the video
THE VIDEO PROCESSING SYSTEM lhl
o- 0 0 0
ó5
4-1
i
O ó
roi
o
E ó
L 5¢
V
.dFF..
C
Ö m
output module. In the E -E mode, the new burst is not generated; therefore,
a burst -notch circuit is provided so that the old burst is not eliminated by
the clamp in the chroma channel. The video at this point is called
"processed video," and it is delayed by 0.5 microsecond in order to match
the delays imposed by the sync channel. The chroma portion of the video
signal is fed through a high -frequency boost circuit which compensates for
any loss that results from signal handling.
Sync Adder
Composite sync from the sync logic module is applied to the sync adder
circuits. Sync is added to the processed video at the output of the video
line drivers. Sync gain may be varied either locally on the front panel of
the vertical advance module or from a remote panel. The composite video
signal is supplied to the picture monitor and waveform monitor (cro) .
Composite sync may be removed from the composite video signal in one or
all video output channels (1, 2, or 3) by opening the sync switches, which
are located in the video output module.
International Standards
Most of the circuitry in the three pulse- forming modules consists of
simple time -constant changes in order to obtain the correct pulse widths
needed for the different tv line standards. The verical advance module
contains the circuitry necessary to change the timing of the advance verti-
cal pulse (3.5H) so that it is compatible with 405- and 625 -line standards.
This circuitry includes an additional transistor stage, a miscount amplifier
for driving the first counter when on 625 -line standards.
The switch used in selecting the different tv line standards is located
directly beneath the picture monitor. For operation on 625 -line standards,
the center frequency of the bandpass filter in the chroma channel and the
band -rejection filter is changed from 3.58 megahertz to 4.43 megahertz by
the action of the line standard switch.
In machines equipped with PAL, when 625 -line high -band color opera-
tion is selected, the regenerated burst switches between 0 and 90° every
other line. The regenerated burst moves back and forth with the meander-
ing vertical interval according to the PAL - system specifications.
5
v a
>
o ._
.-vv\--1 H,,
"lQ.Q.p 1
G°
E C<
>
_ +N//- Ñ
ce
>
N
0
z
The incoming video signal from J008s enters the circuit board at pin R
and is coupled by emitter follower Q6 through resistor R15 to the junction
of diodes X1 and X4. The anodes of these diodes are connected to the
emitters of Q2 and Q5, respectively.
The input impedances of diodes X1 and X4 and transistors Q2 and Q5
are the key to the remote control of video gain. The dynamic impedance
of any semiconductor device is a function of the dc current through it. The
resistive component of the diode impedance is approximately 25 divided
by the dc current in milliamperes. Thus if the dc current is 1 mA, the
resistive component is 25 ohms; if the current is 5 mA, the resistive com-
ponent is 5 ohms.
The impedance at the junction of R15 with the cathodes of X1 and X4
is very low (i.e., of the order of 10 to 20 ohms) . The resistance of R15
(680 ohms) is considerably larger than the junction impedance. It thus
establishes a constant -current source that is relatively independent of any
current changes that occur in later circuitry. Fig. 6 -14 is a simplified
diagram that illustrates the mechanism involved. At the junction of R15
with diodes X1 and X4 there is a variable resistance to the left and a
variable resistance to the right. The constant current from Q6 divides
between the two grounded -base amplifiers (Q2 and Q5) in the inverse ratio
of their dynamic impedances.
The output of the circuit of Q2 and Q5 is taken from the collector of Q2.
When Q2 saturates, Q5 cuts off, and all the dc current through R15 passes
through Q2 and R12. This is the condition for maximum gain. When Q2
cuts off, all the current through R15 passes through Q5 and R8. This is the
no- signal condition.
If transistors Q2 and Q5 were used by themselves and their conductions
varied differentially, the dc voltage at the collector of Q2 ( the output)
would follow the changing dc level, causing an unwanted "bounce" in the
resulting gain. Such a sudden change of dc level would seriously disturb
subsequent clamp circuits in the processing amplifier. Transistors Q3 and
Q4 provide a dc current change which is (ideally) equal and opposite
+12 V
R8
-12 V
THE VIDEO PROCESSING SYSTEM 185
to the "bounce" current that appears in R12. The dc current through R12
remains nearly constant throughout the range of video gain. The bases of
Q3 and Q5 are connected through R14; the collectors of Q4 and Q5 are
common. Thus Q3 and Q4 act to cancel the dc current and tend to hold the
collector of Q2 at a constant dc voltage. (Neither Q3 nor Q4 is concerned
with the video signal.) The network consisting of R4 and C3 (Fig. 6 -13)
provides dc feedback to further reduce the unwanted gain "bounce."
When relay K1 is de- energized, panel control R003B (VIDEO GAIN)
controls the dc bias of Q1 and Q3; when K1 is energized, an identical
potentiometer on the remote control panel replaces R003B.
The range (or calibration) of R003B is set by R18. Control R20 adjusts
the base bias of Q4 to balance the currents through Q3 and Q4, and to
equalize their currents with the currents through Q2 and Q5. An RC net-
work that includes R14 reduces differential phase and differential gain
and maintains the frequency response through etched board 18. Inductor
L1 peaks the frequency response of the collector output of Q2. Inductor L2
isolates feedthrough of the video signal appearing at the collectors of Q2
and Q3. This is done to prevent high- frequency loading of the signal by
the emitter of Q4.
The output of Q2 is coupled by emitter follower Q1 through response -
equilizing control R1 to pin M, from which it is routed to the remote
position of switch S2002Á ( VIDEO OUT) The network composed of R1,
.
carries only dc control currents and voltages, and it may have any reason-
able length.
EXERCISES
Q6 -1. What is the basic signal error in the playback of a tape when the guide
positioning is not identical to that used for recording?
Q6 -2. What is the basic element in any quadruplex -tape ATC system?
Q6 -3. How is the element of Q6 -2 basically used for time -error correction?
Q6 -4. With the tape system in the nonsynchronous mode of playback, what
are the basic signals used for comparison to generate an error signal?
Q6 -5. With the tape system in the synchronous playback mode, what are the
basic signals used for comparison to generate an error signal?
Q6 -6. What is the normal correction range (in nanoseconds) for automatic
time correction circuitry?
Q6 -7. Are the varicaps in an electronically variable delay line forward or re-
verse biased?
Q6 -8. What signals are used for comparison to generate an error signal in
color ATC systems?
CHAPTER
7 -1. RELATIONSHIPS
The basic synchronizing function of the tv -tape system will be clear if
the following basic relationships are clearly visualized:
1. Relationship of the head timing pulse (tachometer) to the video head
recording vertical sync.
2. Phasing of the control -track signal being recorded with the video
tracks being recorded, as timed from the 60 -Hz reference pulse from
which the 30 -Hz frame (edit) pulse is derived.
3. Relationship of the tracking control to the video head playing back
vertical sync.
Fig. 7 -1 illustrates the factor that determines which of the four video
heads records the vertical sync. Since the 240 -Hz signal derived from the
rotating-head shaft is compared in phase with the 240 -Hz signal derived
from vertical sync (reference pulse) , the particular head in contact with
the tape at the leading edge of this pulse is the one that will be recording
vertical sync. Thus, the physical orientation of the timing ring of the tone -
186
THE SERVO SYSTEM 187
Timing Pulse
From Head Speed
J L Timer
ITachometerI
Output Feedback
240 Hz
I
60 -Hz Ref
Pulse Derived
From Vert Sync
160 Hz
x4 -
?40 Hz
Ph se
Comp rison
240-Hz
Osc
wheel notch relative to the head assembly determines which of the four
heads will record vertical sync.
The notch in the RCA tonewheel is directly over the tonewheel head
when video head No. 1 is in the approximate center of the tape. Thus
video head No. 1 records vertical sync in the RCA machine. The orientation
of the Ampex tachometer is such that the leading -edge reference occurs
when head No. 4 is in the approximate center of the tape. Thus, head
No. 4 records vertical sync in the Ampex machine.
The A -scope ( which monitors the playback channels) on the Ampex
control panel is a driven type, with the horizontal sweep being triggered
from a derivative of the timing tachometer. The output of the four video
heads during one revolution of the head drum is displayed on one hori-
zontal sweep of the oscilloscope. With the tracking control centered (same
head playing back vertical sync as recorded it) , the presentation is as shown
in Fig. 7 -2. Since vertical sync occurs at the approximate center of the
channel-4 track, the retrace is triggered at this time, and channel 4 con-
tinues a short distance on the left section of the trace.
NOTE: It should be understood that any existing difference with respect
to which video head records vertical sync does not affect compatibility or
interchangeability of tapes between any two systems. When a recorded
II2
Fig. 7 -2. A -scope presentation of
Ampex head switching.
4
III
2
IIII E i pi IIII IIII
tape is played back on a different head than that used for recording, the
tracking control is adjusted to obtain the best match of head-to -tape
response.
The phasing of the control -track signal, recording of the video tracks,
and frame (edit) pulse is illustrated in Fig. 7 -3. (Since the control -track
signal is used to provide a timing reference during playback for the
capstan servo, it is sometimes called the tracking- control recording.)
The 240 -Hz sine wave on the control track represents the current in the
control -track head during recording, and hence the degree (and polarity)
of magnetization on the tape. The current nodes (zero magnetization)
occur in phase with alternate video -track guard bands. Thus, current nodes
and current maximums occur in phase with alternate video -track guard
bands. The control-track sine -wave current is so phased that the edit pulse
(30 -Hz frame pulse derived from the 60 -Hz reference pulse and occurring
every 33 video tracks) occurs at the point of maximum current in the
guard band between the second and third tracks following the track that
contains vertical sync. This is between tracks 3 and 4 if track 1 contains
the vertical sync, as shown in Fig. 7 -3.
It is important to note that proper phasing of the control -track signal
with the edit pulse (and hence the reference pulse) automatically estab-
lishes a standard phasing relative to the video tracks being recorded; this
is most important for interspliceability. If a tape is spliced onto another
segment which does not have identically the same phasing between video
tracks and control tracks, the tracking control must be manually readjusted
immediately after passage of the splice.
Tape Travel 00
90° 180° 270° 360°
Tape
Head
Edit
VScans
Pu se Tachometer
T
Gen
Video 30 Hz 240 Hz
Tracks
Start of Vert
During
Recording - 240 Hz
Rotating Heads
/
Blanking 30 Hz
On ly
End of Vert
Control Trac\
Head
Blanking Playback
Current Nodes
(Zero Magnetization)
Capstan
.4 -411. Motor
Capstan
Fig. 7 -3. Phasing of video tracks, control -track signal, and edit pulse.
THE SERVO SYSTEM 189
Fig. 7 -4 shows the relationship of the tracking control to the video head
playing back vertical sync. As shown, either the control-track signal or the
head -tachometer signal may be shifted in phase so that any one of the four
video heads will track on vertical sync. Ampex machines track by means of
the control -track signal, and older RCA machines track by means of the
tone -wheel signal. All older RCA systems modified for synchronous opera-
tion and all later-model RCA recorders also use the control -track signal
path for tracking control.
Tape
Variable
240 Hz
' Phase
Older RCA
Rotating Heads
r-
..
. . .. . . __.--Imo . . r
/ Tv:
O
O
O
E
-ó
Ñ
v
m
rá
ó d
vc E -8
c
°
Á
LL
«t. iw
Ó
a ¡
...... __..._
Ó
rTo
/
I . . .. _____
.
rl jv
o
,;§
. . . .
.....
ÿ N
-
. ... ,- -___-......
Fig. 7 -5. Positional relationship of video and sync on coated surface of tape.
1 92 TELEVISION BROADCASTING
r
Drum -
Tach
Drum -lach
Sh per
Dr um-
Phase
I -0--
Drum
Oscillator
Input Comparator o
Electronic Switch
Control -
Record
Control-Track
i
Track Playback Phase Cap tan
Comparator Oscillator
Head Play Amplifier
-
information at the precise center of the tape width.
Incoming
Video Sync-1'o Timing Voltage
Phase
o Error Controlled
Modulator
Mode °o Detector Drum Osc
Selector
Switch
1
Tach Amp 1
Drumer
ad Dr
Sh per Head -
Dr um
Motor
Tach Signal
Output
Phase Drum -Motor
From Drum
Modulator Driver
Oscillator Head -
Drum
Motor
Tachometer
Frequency
Shaper and Tach Signal
Discriminator
Amplifier
During the record mode, the capstan servo system maintains constant
velocity of the tape in its movement past the rotating video head drum.
The frequency of the capstan drive signal is 60 Hz (U.S. system) derived
from an oscillator and locked to the drum-tachometer signal. The capstan
synchronous motor is thus caused to drive the tape at a nominal rate of 15
inches per second (U.S. system) but is electronically locked to the rotation
of the video head drum.
Derivation of the capstan drive signal in this manner, rather than directly
from the power line, prevents power -line phase transients from having an
effect because of the "soft- lock" characteristic of the drum servo loop. In
addition, if the head drum changes speed momentarily to rephase, the cap-
stan changes speed proportionally and thus maintains a constant interval
between the recorded transverse tracks across the width of the tape.
Servo Play -Mode Functions
During the making of a recording, the tape moves longitudinally at the
nominal rate of 15 inches per second, while the video heads rotate at the
rate of 14,400 rpm (US system) The carefully controlled positional
.
relationship between vertical -sync information and the 30 -pps frame pulse
is established at this time, during which the tachometer scans the rotation
of the video head drum and generates one complete cycle of a square wave
during each revolution.
Because the drum turns at 240 rps, the square wave produced by the
tachometer has a frequency of 240 Hz and is in exact phase with the drum
position. The tachometer signal is recorded on the control track at normal
level together with the 30 -pps frame pulse, which is recorded at the level
of tape saturation. The considerable difference between the recorded levels
of the frame pulse and the control track permits positive identification of
both reproduced signals.
THE SERVO SYSTEM 195
Voltage -
Phase Phase
60 PPS Cont oiled
Comparator Modulator
Drum Osc
Drive Signal
Damping Error Signal
Tachometer
Amplifier Tach Signati
and Shaper
Reference
Ph se Drum -Motor
Horizontal-. Modulator Drivers Head -
Sync
Drum
Motor
Tachometer
Frequency
Shaper and Tach Signal
Discriminator
Amplifier
automatic mode, initial condition, and Fig. 7 -9B shows a block diagram of
the automatic mode, final condition.
Track identification is derived from the 30 -pps frame pulse recorded on
the control track, which controls capstan rotation to position the tape
longitudinally. During this period, head -drum rotation is locked to a
reference signal (e.g., station reference sync) , but the video heads are not
yet in contact with the tape. Thus the longitudinal positioning of the tape
is accomplished by relatively large changes of tape speed during a period
when recorded vertical- and horizontal -sync information is not being repro-
duced. Because the control -track record /reproduce head is in contact with
the tape at all times, frame pulses are reproduced as soon as tape movement
begins. It should be remembered that each frame pulse coincides in time
with the vertical -sync information that initiates a new frame (Fig. 7 -10) ,
and that the frame pulse is recorded on the control track, whereas the verti-
cal -sync information is recorded on a transverse video track.
-If11-_ 1111111
Blanking' HH /21- -H-
Y Third Vertical Serration Vertical Sync
After 1 or 3.8 seconds, depending on the type of start -up used, and after
capstan rotation has stabilized or nearly stabilized the longitudinal position
of the tape, the vacuum tape guide brings the tape into contact with the
video heads; the heads then begin reproduction of the recorded video
information. If for any reason the control -track frame pulses are not present
and framing is not accomplished as described above, duplicate framing
information derived from the reproduce video tracks controls the capstan in
the same manner as the control-track frame pulses. When framing is com-
pleted and video is being satisfactorily reproduced, the capstan servo and
head -drum servo change to the same condition as the horizontal -mode final
condition.
Horizontal Mode -In
this mode, capstan rotation is controlled by phase
comparison of the reproduced control-track signal with the tachometer sig-
nal. There is no other correction of the longitudinal position of the tape.
Fig. 7 -11 describes the horizontal mode in the initial condition; Fig. 7 -9B
198 TELEVISION BROADCASTING
describes the horizontal mode in the final condition (same as the automatic
mode in the final condition) .
During the few seconds after the reproduced control -track signal appears
and before the vacuum guide engages the tape with the video heads, head -
drum rotation is stabilized by phase comparison of the tachometer signal
with reference vertical sync. At the conclusion of this interval, if repro-
duced sync pulses are being received, the control of head -drum phase is
transferred to the output of the horizontal comparator. It remains under
this control unless the tachometer phase comparison shows a slipping
condition of the tachometer phase with respect to reference vertical sync
or unless the reproduced sync completely disappears.
Reproduced
a, Frequency
Control Track 1
240 Hz Discriminator
Signal
Voltage Capstan -
Phase Capstan
Comparator Controlled Motor
Motor
Capstan Osc Drivers
.---240 Hz
Phase
Comparator
Voltage
Controlled
Drum Osc
Reference
Frequency
VerticalDiscnmmator
60 Hz
Phase
Modulator
Dr um -Motor
Drivers
Sync Head -
Drum
Motor
Tachometer
Frequency
Amplifier Tach Signal D
Discriminator and Shaper
Voltage
Reproduced Phase
Controlled
Vertical Sync 50/60 pps Comparator
Drum Osc
Frequency
Discriminator
The TRACKING control on the main control panel has the "home" track
setting marked on its dial. At this setting, video head 4 will reproduce the
transverse video tracks that contain the vertical-sync information.
While a given video head assembly is in service, the TRACKING control
should be kept at the "home" track setting during the recording and repro-
duction of a given tape. This will assure the best possible reproduction of
that tape because the reproduce conditions recreate those present during
the making of the recording.
By altering the setting of the TRACKING control, any one of the other
video heads may be selected to reproduce the tracks that contain the vertical -
sync information. During the reproduction of a recording made by a
different video head assembly (or a different system) , there may be an
occasional subjective improvement of reproduced picture quality when
head 1, 2, or 3 is selected to reproduce these particular tracks.
Normal Mode-The normal mode is recommended for monochrome
reproduction when the recording is known to contain sync- timing discon-
tinuities such as those that occur when a tape is improperly spliced. Re-
produced video is "soft- locked" to the power -line frequency. However,
reproduced sync timing is tc ..ally ambiguous with respect to the station
sync -generator reference. Tape reproduction results without regard to
framing and is based on phase comparison of the 240 -Hz control -track
signal with the drum -tachometer signal, which in turn is phase locked to
the power line. Fig. 7 -13 describes the Intersync servo when operated in
the normal mode.
Reproduced
Control -
Frequency
Track 1 r
Discriminator (Connected Only Before
Signal 1
Capstan
Phase Capstan Capstan
Motor
Comparator Oscillator Motor
Driver
I
Ph se Drum Phase Drum Motor
Tach Signal
Comparator Oscillator Modulator Drivers Head -
Drum
(Connected Only Motor
Before Phase Lock
Power f Is Achieved)
Frequency
Line Tach Signal
60 Hz
Discriminator Damping Error Signal
Reference
Tachometer
Frequency
Amplifier
Discriminator
and Shaper
Preset Mode -In the preset mode, the controls of the Intersync servo
unit may be set to any desired record and /or reproduce reference. This
mode is used mainly for checking various areas of the circuitry and to
meet unusual requirements of the user. The preset mode is similar to the
normal mode except that its timing reference source is chosen by means
of two chassis- mounted selector switches labeled PRESET REFERENCE
SELECTORS.
d
3 óÓ
CO
CU ai
ó
Uco
5 ó
0
s_
3 - 3
o - ó0 c ú
E n.o E d E =
Y N U
ó á o á
U
U Ó
> O U -==
ú00
_ Ff
CC
,..
_
>
ú
c
ó
m n
N
c O
CC (.D
NT
tape sync and the error buses of the monochrome or color ATC system
(Chapter 6) .
Record Mode
In the record mode, an afc circuit in the tonewheel processor module,
locked to the pulses from the magnetic tonewheel attached to the headwheel
motor shaft, produces a sine wave at twice the tonewheel frequency. This
sine wave serves as the basic motor drive signal (see Fig. 7 -15) Networks
.
in the headwheel modulator module split this signal into three sine waves
which are 120° apart in phase. Chopper circuits controlled by a dc error
signal from the tonewheel servo module then pulse -width modulate the
three sine waves at a line- frequency (horizontal) rate. The amplitude and
polarity of the dc error signal control the modulating pulse width, and,
since the drive -signal amplitude is zero during the pulse interval, the pulse
width determines the energy remaining in each cycle of the sinusoidal
signal after modulation. For example, a negative error signal of maximum
amplitude reduces the pulse width to zero and leaves the sine wave in the
unmodulated, or maximum- energy, condition. Conversely, a positive error
signal of maximum amplitude increases the pulse width to a full horizontal
line period, thereby leaving zero energy. Filters in the modulator output
remove the line -rate information and leave three sine waves, separated in
phase by 120 °, with amplitudes depending on the energy in the pulse -
modulated signals. Separate power amplifiers (modules 606, 607, and 608)
are driven by each of the three headwheel- modulator output signals and
provide just enough power to drive the three -phase synchronous motor at
one -half the synchronous speed.
As the headwheel rotates, the tonewheel pickup head provides one pulse
per headwheel revolution. These pulses provide feedback to a velocity -error
detector and a phase -error detector in the tonewheel servo module. The
velocity -error detector compares the period of the tonewheel pulses with
an independent fixed reference, which is the combined period of two free -
running multivibrators. The total multivibrator period is one-quarter of
204 TELEVISION BROADCASTING
-i
3
n
ao E E
ñ
, E
x,f
N
Q
J 2ó
O O
u
Ó
ú
r o
O
L n
pV a_
Ó
C
U
CL
y u ó
> ÿ CO J
CC N 0 c 2á c
$
8 i ó cá S
p
Op m
C U
V1 J U
_11¿ N CL
E
c'D
ÑCe N
Ho
u
c
O
W
L
oC
L
ry
_ - J
u
VI
Ñ
--a-.1
Ñ
d 3
cd
_
p
1.
N
Ñ
vÑ
>
Só
m ó
n
ó
Fig. 7 -15. Headwheel servo system during recording mode and tonewheel or
Switchlock playback mode.
THE SERVO SYSTEM 205
too rapidly for the capstan servo. Therefore, to permit a smooth transition,
a special circuit designated the Lazy Servo Simulator ( LASS ) temporarily
reduces the phase gain after a nonsynchronous switch ( see Fig. 7 -15) The
.
power -line signal may be selected for timing the vertical reference pulse.
(The timing signal is obtained by utilizing a switch on the reference
generator module and, if a power -line signal is desired, by changing a
jumper internally.) This mode of servo operation will produce a satis-
factory picture, but tape and reference sync will not necessarily be aligned
because of the absence of feedback from the demodulated (tape) video
signal to the headwheel or capstan servo.
Switchlock-When Switchlock is selected, the headwheel operates in
the tonewheel mode of servo operation, but the capstan servo controls the
tape motion so that the headwheel is on the track containing recorded
vertical sync at the end of a frame when the corresponding vertical inter-
val of reference (local) sync occurs. This coarse framing brings tape and
reference vertical sync close enough to permit roll -free switching be-
tween the recorded output and other video sources, or to permit electronic
splicing; however, it does not correct for errors in placement of the re-
corded vertical sync on the tracks.
NOTE: Before an electronic- splice recording can be made, the servo selec-
tor switch on the tape sync processor module must be in the SL position,
and the machine must be actually playing back tape in the Switchlock
mode of servo operation. If this condition is not met, the splice control
circuit automatically prevents entry into the splice record mode.
Pixlock operation differs from Switchlock only in the use of error infor-
mation from the Linelock module in addition to the velocity and phase
information provided by the tonewheel servo module. The combined oper-
ation of the capstan and headwheel servos in establishing Pixlock occurs
in two main phases: the tonewheel -tva phase and the Linelock phase.
At the moment the servo mode -selector push- button switch on the play
panel is pressed for Pixlock operation, the machine enters the tonewheel-
tva phase (Fig. 7 -16) In this phase, a variable delay generator (time -
.
larger gain not attainable at this low sampling rate. While tape and refer-
THE SERVO SYSTEM 209
ence vertical signals are being brought into alignment, a coincidence de-
tector and lock sense circuit in the Linelock module, independent of the
tva circuit, compare the tape and reference vertical phasing. When verti-
cal coincidence is established, this circuit energizes a bus (lock bus) which
causes relays to switch the headwheel servo to the Pixlock mode.
At the moment the system is switched to the Pixlock mode of operation,
the tva error signal is disconnected, and the Linelock error -detector cir-
cuitry is enabled (refer to Fig. 7 -17). The Linelock error detector com-
pares an output derived from tape horizontal sync at the video demodu-
lator output with a reference derived from local ( house) horizontal sync,
to produce an error signal sampled at the line rate. Filters and amplifiers
separate this signal into a high- frequency, or velocity, ac component and
a low- frequency, or phase, dc component. The high- and low- frequency
error- signal paths have separate antihunt and stabilizing networks. A gain
control in the low- frequency error- signal path permits obtaining optimum
loop gain and phase characteristics. In addition, each output amplifier pro-
vides special dynamic characteristics which compensate for the mechanical
response of the headwheel motor to sudden disturbances.
The low- frequency Linelock error component is fed to the time -modu-
lated vertical circuit in place of the tva error, and the high- frequency
Linelock error component is fed, through a separate path, directly to the
headwheel modulator. When the Linelock error -detector circuit is en-
abled, a time -delay circuit gradually increases the gain to prevent tran-
sients from supplying false information. As the gain increases, the low -
frequency error component assumes control of the time -modulated verti-
cal circuit; thereby it adjusts the delay previously established by the tva
error signal to a value which causes the tonewheel servo to align tape
horizontal with reference horizontal. Simultaneously, the high- frequency
Linelock error signal provides velocity information to the modulator
which locks the frequency of tape horizontal sync to that of reference
horizontal sync.
Completion of all phases of Pixlock operation requires less than five
seconds from the moment of entry into the play mode. When the Line -
lock phase is completed, only the mechanical capabilities of the motor
drive system limit the tightness of control. The high gain and rapid re-
sponse of the combined tonewheel and Linelock system reduce the maxi-
mum phase error between tape and reference sync to less than ±0.07
microsecond for a tape recorded on a machine having the same type of
servo system.
As a distinguishing characteristic of Pixlock operation, the vertical -
coincidence circuit continues to check the phase relationship between tape
and reference vertical sync after the machine enters the Linelock phase.
Therefore, if dropouts or similar disturbances cause loss of vertical coinci-
dence, the lock -sense circuit will temporarily switch the headwheel servo
back to the tonewheel -tva phase until the tva circuit re-establishes coinci-
THE SERVO SYSTEM 211
Play Bus
Platch Bus to
Lock Bus Vertical -Coincidence
From Circuit
Linelock
Circuit Bus Asking For
Linelock Servo
Platch bus then disables the vertical- coincidence circuit of the headwheel
servo, and a relay switches the capstan -servo reference from the local frame
pulse to a frame pulse derived from the tonewheel. However, if Pixlock
is not achieved within eight seconds, the Platch circuit is automatically
energized, and the Platch bus disables the vertical- coincidence circuit. The
lock -sense circuit then switches the system to the lock phase.
The Platch bus thereafter remains energized throughout the playback
period. The Linelock circuitry then functions exactly as described above
under Pixlock, except that a loss of vertical framing does not cause rever-
sion to the tonewheel -tva phase. After a disturbance, the Linelock circuitry,
by rephasing the headwheel, locks the nearest horizontal line of tape video
to reference (local) sync, and the capstan servo, referenced to the tone -
wheel, maintains tracking by adjusting tape motion.
Also, throughout the discussion of the capstan servo system, the term
"frame pulse" is used frequently. In the tape recording and playback
processes, there are actually several distinct types of pulses occurring at
a frame rate (30 Hz on domestic standards or 25 Hz on international
standards). One type, designated the reference frame pulse, is obtained by
comparing station (reference) horizontal- and vertical-sync signals in a
coincidence gate circuit. During playback, this pulse is used directly as a
reference for the capstan servo. During recording, the pulse is superim-
posed on the control -track signal, and the combined signal is recorded on
the tape as specified by the SMPTE recommended practices. The frame -
pulse portion of this combined control -track signal is referred to either
as an edit pulse or as a control -track frame pulse. During tape editing, this
pulse serves as a frame marker on the tape to permit making roll -free
mechanical splices. Also, during the first four seconds of playback, the edit
pulse is separated from the combined signal picked up by the control -
track head and used in the fast lock-up circuit of the capstan servo.
In addition to the reference and control -track frame pulses, a tape frame
pulse is used during playback in the Switchlock and Pixlock modes of
servo operation. This pulse is generated by feeding tape horizontal and
vertical sync (separated, after demodulation, from the signal picked up by
the video heads) to a coincidence detector. The tape frame pulse, there-
fore, occurs at a definite time during the vertical -blanking interval at the
end of each frame of the tape video signal.
The binary counters produce a 60 -Hz square -wave signal which is con-
verted into a sinusoidal signal and fed to a power amplifier. The power
amplifier then provides the power required to drive the capstan motor
(Fig. 7 -19).
To provide a continuous, accurate record of the capstan motor speed
during recording, a 240 -Hz sinusoidal control -track signal is recorded
along the lower edge of the tape by the control -track head located on the
headwheel panel. (The control track is utilized in controlling tape motion
during playback and is analogous to the sprocket holes in motion -picture
film.) The control-track signal is derived from the tonewheel pulse signal,
which passes through a variable delay circuit before being formed into a
sinusoidal waveform, and thus bears a definite phase relationship to the
video heads. (It should be noted that when the recorded control -track
signal is played back, it will not appear to be sinusoidal; because a direct
recording process instead of a linear recording process is used, the signal
will exhibit a characteristic double- humped pattern. This is clarified in
Section 7-5.)
Superimposed on the control -track signal is a stable timing reference
pulse designated the control -track frame pulse. This pulse is produced by
a coincidence gate circuit which compares reference vertical- and refer-
ence horizontal -sync signals. The pulse occurs at a 30 -Hz rate (25 -Hz rate
in international machines) , and thus there will be one control -track frame
pulse for every eight cycles of control -track signal (every ten cycles in
international machines) To meet SMPTE standards with regard to rela-
.
tive phasing between the control -track signal and the control -track frame
pulse, a variable delay is incorporated into the circuit which generates the
control -track signal. Thus, by maintaining a specific phase relationship, the
control -track frame pulses provide physical locations on the tape (approxi-
mately every half inch) at which the tape may be cut and spliced without
danger of picture rollover (loss of vertical sync) during tape playback.
Since the control -track frame pulse may be utilized as a guide for cutting
the tape during tape editing, this pulse is sometimes referred to as an
"edit" pulse.
The level of the combined control-track signal and control -track frame
pulse fed to the control -track head is fairly critical. Therefore, the simul-
taneous playback head is provided as a means of playing back the com-
bined signal immediately after it has been recorded, so that the signal
level may be observed on the cro monitor and corrected if necessary. The
combined signal may also be observed directly on the monitor as it is be-
ing recorded. This allows the operator to position the control -track signal
relative to the control -track frame pulse precisely and to obtain the cor-
rect pulse amplitude. In addition to these presentations on the cro moni-
tor, provision is made for observation of the motor drive signal as a
means of checking the normal output which drives the capstan power am-
plifier during the recording process.
216 TELEVISION BROADCASTING
r 7
C N
.0 2N iao
ó V
0
o >
L. N
u--
La, H
N LN ó 0,.°.
ÿ úo .p vN C *C
o {n U o
ti 0_ S
RI
uOiIOW adel
Playback Operation
During playback, several modes of operation are possible. These are
described as follows.
Tonewheel-During tape playback in the tonewheel mode of servo
operation, the capstan servo controls the longitudinal tape speed and phas-
ing so that the video heads track properly on the recorded video tracks.
In playback operation, the capstan motor speed is controlled by a local
oscillator rather than by the tonewheel pulse as it is during the recording
process. (See Fig. 7 -20.) The oscillator is "free- running" at a frequency
of 240 Hz; however, if the tape moves too slowly or too rapidly, an error
signal is developed which increases or decreases the oscillator frequency
(and thus the capstan motor speed) to correct the tape speed. The 240 -Hz
output signal from the oscillator is divided by four by the binary counters
and formed into a 60 -Hz capstan -motor drive signal, in exactly the same
manner as is the tonewheel pulse during recording.
Run 9
Edit Pulse
d0 Start
4 -Sec Swi ch
/Control-Track
\RecIPB Head
-r
Con trol -Track
Pulse Generator
(Control Track
RecIPB 6171
111.1
Variable Delay
(Capstan Phase
6161
- CT
:8 Counters
1:10
Inter nation all
(Capstan Phase
6161
Control Voltage)
240 -Hz
Binary
Oscillator
Power Counters
(250-Hz
Capstan Amplifier
605
e4
-- I nternational1
(Capstan (Capstan
Oscillator 6141 Oscillator 6141
The error signal which modifies the local -oscillator frequency to correct
the tape speed is generated from the phase comparison of a pulse derived
from the playback control -track signal and a trapezoid waveform derived
from the reference frame pulse, which in turn is derived from local (refer-
ence) sync. The 240 -Hz control -track signal is played back from the tape
by the control -track record /playback head and is amplified, limited, and
formed into a series of 240 -Hz pulses. The 240 -Hz pulse frequency is
218 TELEVISION BROADCASTING
These pulses sample the slope of the reference trapezoid waveform to ob-
tain the error signal. Local ( reference) horizontal- and vertical -sync sig-
nals are combined in a coincidence gate circuit to form pulses at the frame
rate, and the reference frame pulses thus formed generate the reference
trapezoid waveform. Therefore, as the tape speed increases or decreases,
the control -track pulse frequency follows this change, and the error signal
generated causes the capstan motor speed to increase or decrease until the
divided frequency of the control -track pulses coincides with that of the
reference frame pulses ( i.e., the capstan motor speed is locked to the ref-
erence pulse frequency) .
During the first four seconds after the PLAY push button is pressed, con-
trol -track frame pulses (edit pulses) if present on the control track, are
,
used to reset the binary counters which divide the frequency of the 240 -
Hz control -track signal. This permits achieving coarse framing (approxi-
mate alignment of tape vertical sync with local vertical sync) before the
vacuum -guide air solenoid is activated and video is obtained, since the
control -track head is already in contact with the tape. Application of the
edit pulses to the binary counters is controlled by the fast lock -up capstan
circuit, which consists of a relay controlled by multivibrators. Although
this circuit is active during the first four seconds of playback regardless
of the mode of servo operation, it is primarily useful in the Switchlock,
Pixlock, and Linelock modes. In the tonewheel mode, after the relay dis-
connects the edit pulses at the end of the four -second interval, the binary
counters are no longer reset. Therefore, if framing is lost because of a
disturbance, it will not be regained, although the frequency- divided con-
trol -track pulses will relock with the reference frame pulses. This is con-
sidered normal for the tonewheel mode, since framing is not a requirement.
Provision is made for continuous manual variation of control -track
phase over a very wide range. Manual phasing is accomplished by varying
the CT PHASE control on the play control panel, which in turn varies a
delay inserted into the control -track pulse path. (The CT PHASE control has
two ranges: a wide range for coarse phasing and a much narrower range
for fine phasing.) The basic purpose of the manual delay is to provide a
means of precisely centering the video heads over the recorded video
tracks. In the tonewheel mode of servo operation, the CT PHASE control
may also be utilized in "slipping tracks" so that video head 1 will play
back information from the video track containing vertical sync (which is
normal operating procedure in the RCA system) or from any of the
three remaining tracks. Provision is also made for manual control of the
capstan motor speed, and thereby of the tape speed, during playback by
interrupting the error signal fed to the 240 -Hz local oscillator and insert-
ing a constant dc current. The constant dc current is inserted by pressing
the red (to speed up the tape) or black (to slow down the tape) CAP-
THE SERVO SYSTEM 219
STAN SPEED button on the play control panel. This feature is useful for
synchronizing two machines which are playing back the same information.
In the Switchlock and Pixlock modes of servo operation, the capstan
servo operates as in the tonewheel mode except that when the edit pulses
are disconnected from the binary counters at the end of the initial four-
second interval, the edit pulses are replaced by tape frame pulses (Fig.
7 -21) . (If no edit pulses are on the tape, or if they are of poor quality,
tape frame pulses are fed to the counters as soon as they are available.)
Use of the tape frame pulses for resetting the counters ensures that the
headwheel will be on the track containing vertical sync at the end of a
recorded frame when the corresponding vertical interval of local sync
occurs. This degree of control, designated coarse framing, or Switchlock,
brings tape and local (reference) vertical sync close enough together to
permit roll -free switching between the tape- machine output and another
video source, or to permit electronic splicing. (In the Pixlock mode of
servo operation, a finer degree of framing, which corrects for errors in
Headwheel
622
(Part of
Video Proc) Edit Pulse
c
- (Capstan Phase
6161
Reset Pulse
1
Coincidence Sens
CT
Control -Track
Variable Delay 8 Counters
Pulse Generator 1:10
Control -Track
RecIPB Head International)
(Control Track (Capstan Phase
( Capsta Phase
Rec/PB 617) 6161
6161
Tape
Reference Reference Reference Comp rator
Vertical Sync Frame -Pulse Trapezoid Bridge
Generator Generator
Control Voltage
240 -Hz
Binary
Oscillator
Power Counters
1250 -Hz
Capstan Amplifier 4
International)
605 (Capstan (Capstan
Oscillator 614) Oscillator 614)
Fig. 7 21. Capstan servo system for playback in Switchlock or Pixlock mode.
220 TELEVISION BROADCASTING
Magnetic
240 Hz
HeadwherI
¡ Tonewheel i 218 +8 Pulse
(250 Hz s8 Counters
i -
'
Tonewheel Servo
International) Processed Tonewheel Pulse UID - I Linelock
International)
(Tonewheel (Tonewheel i Norm
Processor 6121 Servo 6111 i
(Control Track
-r
Variable Delay
(Capstan Phase
- s 8 Counters
1:10
In ter n utionall
(Capstan Phase
Horizontal
Sync
Vertical
Sync
f
Control Voltage) Were ce Trapezoid
240-Hz
f-
Binary
Oscillator
Power Counters (250 -Hz
Capstan Amplifier a4
International)
605 (Capstan (Capstan
Oscillator 6141 Oscillata 614)
to the control -track pulse counters in the capstan servo system, are reset
by the reference frame pulses so that the divide -by -eight pulse coincides
with the tonewheel pulse closest to the reference frame pulse. This elimi-
nates picture disturbances during the transition period after the capstan -
servo reference is switched from reference frame pulses to tonewheel
divide -by -eight pulses.
When the headwheel phasing changes, the tonewheel -pulse counting
speed will vary depending on the direction of change, but the counter
output will remain properly phased with respect to the reference frame
pulse. If the headwheel speed increases, the counters will count tonewheel
pulses faster, and the capstan motor speed will increase until the frequency -
divided control -track pulses are again locked to the frequency- divided tone -
wheel pulses.
Splice Modes -In
machines equipped with the accessory electronic splic-
ing feature (Chapter 10) , the capstan servo system operates as previously
described except when an insert or add -on splice recording is being made.
(In an insert splice recording, the video on a section of a previously re-
corded tape is erased and new material is recorded in its place, leaving
portions of the originally recorded video immediately preceding and fol-
lowing the newly recorded section. In an add -on splice recording, however,
the new material is added at the end of the original recording.) As far as
?2 TELEVISION BROADCASTING
the capstan servo system is concerned, the essential difference between the
insert and add -on mode of splice operation is as follows: In the insert
mode, the capstan servo system operates as during normal tape playback,
and the previously recorded control track is not erased but is allowed to
remain on the tape. In the add -on mode, the capstan servo system operates
as during normal tape recording, and the original control track (if any ) is
erased so that a new control track may be recorded.
In either type of splice recording, the capstan servo must be controlled
so that no sudden change in tape speed occurs at the junction of the old
and new recordings, and the tape speed during recording will be the same
as during playback. These results cannot be obtained by driving the cap-
stan motor with a signal derived from the frequency- divided tonewheel
pulses, as in normal recording, since the tonewheel pulses provide no indi-
cation of the actual tape speed present when the old recording was made.
Therefore, during a splice recording, the capstan is driven, as in playback,
with a signal obtained by dividing the frequency of the output from the
240 -Hz oscillator. The method of controlling the oscillator frequency de-
pends on whether an insert or add -on splice recording is being made.
For an insert recording, the control -track signal previously recorded (at
the same time as the old video signal) is already present along the entire
length of tape, and this signal contains precisely the desired information
regarding the speed of the old recording. Therefore, instead of erasing the
old control track and recording a new one as the new video information
is being recorded, the old control track is retained and used, as in Switch -
lock operation, to derive an error signal which controls the frequency of
the oscillator. Operation of the entire capstan servo system during insert
splice recording is actually identical to Switchlock playback except that
the reset -pulse gating circuits for the control -track binary dividers are dis-
abled so that the binary counters cannot be reset by noise pulses during
the record interval of the electronic splicing process.
In an add -on splice recording, the section of tape on which the new
video recording is to be made may not already contain a recorded control -
track signal, since this section begins at the end of the old recording.
Therefore, a new control track must be recorded, as in normal record
operation. During the transition from playback to recording, the fre-
quency of the capstan oscillator ( and thus the capstan motor speed) is
controlled by an adjustable dc voltage from a capstan memory circuit, in-
stead of by an error signal. The dc voltage is related to the original re-
cording speed by adjusting a potentiometer on the splice logic module
while playing back the originally recorded section of the tape before start-
ing the add -on recording. For testing the potentiometer adjustment, a
momentary -contact toggle switch on the splice logic module controls a
relay which permits opening the capstan servo loop and temporarily dele-
gating control of the oscillator to the capstan memory circuit while the
machine is in the splice play mode.
THE SERVO SYSTEM 223
Tape
Guide
CR0 Feed
(A) Head functions.
Rec Edit
Spurious
Response
Edit Pulse
(Every 8th Wave)
Record Signal
Ir
-L
Max Record Flux Change
(B) Waveforms.
Fig. 7 -23. Control -track record and playback waveforms.
THE SERVO SYSTEM 225
any of the modes in which the headwheel is running, i.e., during record,
set-up, wind, standby, and any one of the playback modes.
When the machine is first turned on, the circuit is inoperative. Placing
the machine in any of the above run modes causes the HW UNLOCK lamp
to light, indicating that the headwheel has not yet come up to the proper
speed. Once the headwheel achieves correct speed and is locked in, the
indicator lamp will go out. Should the headwheel go out of lock for more
than 0.1 second, the HW UNLOCK lamp will light again.
A logic block diagram of the Hw UNLOCK indicator circuit is shown in
Fig. 7 -24. It illustrates the conditions required to extinguish the light.
Wherever AND and OR gates are shown, a small circle at the input or out-
put means that a negative (or low) input or output will satisfy the con-
226 TELEVISION BROADCASTING
ti
-8-
2Ó
Q
3
X
ÿU -O.
cro U
>
CC o
H
o óá
Ea
s
1
,á
wa
Eg
6
-
ditions of the gate. Lack of a small circle indicates that a positive (high
ground potential, usually ) input or output level will satisfy the conditions
of the gate.
Operation of the basic circuit is as follows: Tonewheel (TW) pulses
from the tonewheel processor module, and tonewheel phase -sample pulses
from the reference generator module are fed through emitter followers
Q22 and Q23 to transistors Q19 and Q21. These transistors form a coin-
cidence gate and are normally conducting heavily. With the arrival of a
negative -going TW pulse and a negative -going phase - sample pulse in
coincidence, the gate is opened, and a positive -going output pulse is ob-
tained at the collectors. This pulse, differentiated and clipped, is fed as a
trigger to monostable multivibrator Q17 -Q18. The positive -going square
wave at the collector of Q17 is applied to an integrator emitter follower
stage, Q16. The resulting integrated waveform fed to the base of lamp
driver Q15 is sufficient to cut off Q15 (which is normally saturated) and
prevent the HW UNLOCK indicator lamp from lighting.
A loss of coincidence between the TW pulses and the phase -sample
pulses would cause no trigger to be fed to the base of transistor Q18, and
no square wave would be fed from the multivibrator. The charge on ca-
pacitor C15 would fall (go more negative), transistor Q16 would con-
duct, providing bias current for Q15, and the light would be switched on.
However, transistor Q17 also functions as an AND gate and Q15 as an OR
gate. A discussion of the conditions necessary to fulfill the requirements
of these gates is given below.
As mentioned above, the logic block diagram illustrates the conditions
required to extinguish the light. In the stop mode, it will be extinguished
because the run -bus input to the Q15 OR gate will be low. It will be out
in any run mode because the other input to the Q15 OR gate will go high
when the inputs to the Q17 AND gate have been satisfied by the appro-
priate lock conditions.
In all run modes (with one exception to be described later) , coinci-
dence of the TW pulse and the phase- sample pulse is required to provide
a high level to input 1 of the Q17 AND gate. Coincidence of these pulses
is indicative of tonewheel lock and is a necessary, but not sufficient, con-
dition to assure that all the requirements of a particular servo mode have
been met. Input 2 to the Q17 AND gate supplies the remaining condi-
tions, involving somewhat more complex logic, as shown in the diagram.
These conditions are described by modes below.
logic driving the lock bus is contained within the servo modules and
requires that the Pixlock playback mode be selected and vertical coinci-
dence be achieved.
Motion
ti
Guide
/
Guide Position Determines
Tape Stretch
Er or
Detector Amplifier Guide
Potentiometer
Drive Motor
(Module 6211 (Module 5061
Tape
wide
07 Atmospheric Air
Pressure Holds
Tape Motion
Courtesy RCA
IIIIIIIJIIIIIIII
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1
A
Horiz -Sync Pulses
Produced by Head 1
I10I
2
I
3 4 5
Horiz -Sync Pulses
Produced by Head 2
OEnlarged Views of
i
Portions of (A)
Timing of pulses
coincides with
II
I
I
L
i
Time D'scontinuity Resulting From
Inadequate Tip Protrusion (Shoe Too Far Out)
average near
OEnlarged Views of
center of head
scanning 114=1. Portions of (B)
interval. II
Enlarged Views of
F Portions of ICI
slight errors in the physical placement of the video heads around the cir-
cumference of the headwheel.)
Movement of the vacuum guide is controlled by a mechanical linkage
from a two -phase induction motor. In the play or record mode, the refer-
ence winding of the motor is continuously excited by the power -line volt-
age, and the control winding is driven by an error signal. Therefore, the
motor shaft normally remains stationary, and it develops torque in one
direction or the other only in response to an error signal. ( The direction
of the torque, and thus the direction of the motor -shaft rotation, is de-
termined by the phase relationship between the 50- or 60 -Hz voltages ap-
plied to the reference and control windings.) The error signal may be
developed manually (during tape playback or prior to recording) or
automatically (during playback only) If the guide position is controlled
.
as a stop against which the guide control arm rests when the guide solenoid
is activated. The extension of the lead screw therefore determines the posi-
tion of the vacuum guide and thus the pressure exerted on the tape by the
video heads. The guide solenoid is actually a pneumatically operated pis-
ton activated by an air valve which is in turn controlled by a relay. The
air -valve relay is driven by a signal from the vacuum -guide diode matrix
board. The vacuum -guide solenoid logic circuit requires that both a head-
wheel velocity lock and a play or record command be present before the
solenoid driver permits the guide to engage. Therefore, the vacuum guide
will engage only when the play or record push button has been pressed
and the headwheel motor has come up to speed (as evidenced by the
velocity -lock -sense control -bus potential developed in the tonewheel servo
module) Thus, it may be seen that if the machine is operated in the set -up,
.
the internal reference module which detects an afc lock. The output from
this circuit controls a relay which provides a ground potential when the
afc lock is attained. The ground potential, fed to the guide servo module,
allows the guide to operate normally in the automatic mode. However, if
an afc lock has not been attained or is lost during playback, ground is re-
moved, and the guide automatically reverts to a manually preset safe posi-
tion. This precaution is important, since an erratic error signal may cause
excessive tape pressure against the headwheel.
To prevent the guide from "hunting," or oscillating, during normal
(nonsplice) recording, the guide motor is disabled following an initial
delay of several seconds that allows the guide to reach its proper position.
If the machine is equipped for electronic splicing, during the splicing
operation the guide motor is disabled immediately when the RECORD but-
ton is pressed; thus, there will be no change in guide position when the
switch from splice play to splice record is made.
In both the Ampex and RCA vacuum -guide servo systems, the circuits
represent an adaptation of a standard servo system designed for a variety
of positioning applications. The main output device for the system is a
motor that normally remains stationary and moves only slightly in one
direction or the other in response to error signals. In general, the motor
automatically seeks to position itself so as to keep the error signal at zero.
The error -detector circuit tuned to 960 Hz (Fig. 7 -25) is used to sense
the presence of timing variations at the head -switching rate. The polarity
of the 960 -Hz signal developed in this circuit indicates the direction of
the timing errors. In the circuitry of the guide -servo chassis, the 960 -Hz
signal is amplified, clipped, and rectified. To preserve the direction sense
in the error signal, the rectifier is a bidirectional type, utilizing 960 -Hz
clamping pulses derived from the tonewheel attached to the headwheel
motor. The clipping is severe enough that if there is any significant error
signal at all, the rectifier produces a fixed dc level of a polarity indicating
60 -Hz Chopper
Motor Winding
60 Hz
+24 V
-24 V
Follow Potentiometer
(Rotated by Motor Shaft)
the direction of the error. The rectifier error signal is converted to 60 -Hz
ac by the action of a chopper -type modulator (Fig. 7 -29) and then ampli-
fied to the power level necessary to drive the guide- positioning motor.
The motor is a two -phase type in which one phase is continuously ex-
cited by the power line, while the other is driven by the amplifier. The
motor develops torque only when a signal is delivered by the amplifier;
the direction of this torque depends on the polarity of the amplifier signal.
The motor then drives the guide through an appropriate mechanical
coupling. The motor movement is limited to only a few revolutions, and
the maximum displacement required for the guide is only about 5 mils.
The servo loop is completed through the video heads, slip rings, video -
playback system, and a sync separator (Fig. 7 -25) .
A much smaller servo loop is used for manual positioning of the guide
during recording operations and during special cases of playback opera-
tion. For manual operation, a potentiometer on the motor shaft is com-
pared with the setting of a second potentiometer, which is mounted on
the control panel (Fig. 7 -29) The motor turns until the voltage at the
.
EXERCISES
Q7 -1. Does the same head which records the control -track signal play back
this signal in the playback mode?
Q7 -2. How is the proper amplitude of the control -track recording signal
determined?
Q7 -3. Name the simulplay heads and give their functions.
Q7 -4. How do you control which head records the vertical -sync interval?
Q7 -5. How do you control which head plays back the vertical -sync interval?
Q7 -6. Give the sequence of headwheel (drum) lock-up in going to the fully
synchronous playback mode of operation.
Q7 -7. (A) Give the sequence of capstan lock -up in going to the fully syn-
chronous playback mode of operation. (B) What requirement for a
synchronous operating mode is dropped when phasing is not required
for special effects or mixing? (C) Why must the capstan reference be
changed when going from A to B?
Q7 -8. Define: (A) Edit pulse. (B) Control -track frame pulse, (C) Refer-
ence frame pulse. (D) Tape frame pulse.
Q7 -9. When it is not required to mix other local signals with the tape signal
output, is it necessary to have the servos under control of the local sync
generator for color tapes?
Q7 -10. In most recent tape systems, why is a trapezoidal waveform generally
used instead of a sawtooth waveform?
CHAPTER 8
color as viewed on the monitor. This is because the color -subcarrier signal
frequency (3.579545 MHz) is exactly 2271/2 times the horizontal line
rate. Therefore, the color information is changed in phase 180° on each
successive horizontal scan line. This makes it necessary to separate the
luminance and color information, because the color signal must be phase
inverted to correct for the color error introduced by the one -line delay.
the sample -pulse eliminator. This is installed in the ATC error detector
module (623) .
RI 21i
...1.-
E-E FM
From R2
FM Ref
508 FM 16 Sensitiv ty
TP3 Set
Ref Frame From
Cont Track
27
Rec/Play
617
Sim DO
TP2
9 I
Cancel From
6
PfaylRec
Panel
Sìmmi
Dropout
Insert 7
Lead Edge Gate +20V
Horiz
From
28 Gate
Int Ref On Voltage
621
32
TP1
30
31 17 Dropout Processor 619
DO Pulse DO FM
NC PIL Bus Play Bus
16181 16181
FM FM In From
19 17
Playback Amplifiers
17
Video FM To Sync FM To
FM Filter Demodulator
15241
DO FM 20 Dropout 17 Video In
Compensator From
DO Det
620 Video Filter
DOC 15 18 527
Cont Video Switch 18 Pio Video
32
27 Video
DOC Out
Video DOC
29 Function
Interlock
Switch
63 Vs
D lay Li e
Prior to
DO Sta d
10
DOC
Interlock Video Pix Video
M2 Driver Out To
Pair 31 MATC Video
624
Tape Sync Proc
622 20
T pe Vert Split
16
a
31A_ 17
14
Sync Post
Charm I
Emphasis
Tape Vert Ref Vert From
Video 529
Ref Gen 613
6
DO 28 ¡20 4 5
Vert ps Del y
2 31
Sync Sync Channel
Coin Gate And Filt r Demod "*-Tape FM I n
1A3 530 17
1 30 From FM Switch
Linelock 522
610
Sync DO Pulse
27
Sample Sample
AND
Pulse Pulse
Gen Elim Gate
ATC E ra Detector
17 623 26
L
The filtered fm signal is applied to the agc circuit across two field -effect
transistors, and it is coupled by an emitter follower to the first of two
feedback amplifiers.
The feedback amplifiers provide a gain of 36 dB. The field -effect tran-
sistors act as electronically variable attenuators in the path of the fm
signal. The field -effect transistors are controlled by the dc amplifier, and
their impedance is inversely proportional to the magnitude of the recti-
fied feedback signal. If the amplitude of the fm signal tends to increase,
the impedance of the field -effect transistors is lessened, and more of the
fm signal is attenuated. Conversely, a reduction in the amplitude of the
fm signal results in an increase in the impedance of the field -effect tran-
sistors, which in turn results in less attenuation. Consequently, the level
of the fm signal being fed to the Dropout Compensator module remains
constant, despite fm level variations such as those encountered between
different fm standards or tape recordings. Amplitude control of the fm
signal is such that the output of the agc circuit remains within ± 1 dB
with an input change of -!-10 dB. Since the agc circuit responds only to
very slow changes in the amplitude of the fm signal, dropouts do not
affect it.
The dropout pulse, derived when a dropout is detected in the fm signal,
is at least seven microseconds long; it may be longer, depending on the
duration of the fm dropout. On being applied to the dropout processor
module, the dropout pulse may follow either of two circuit paths, depend-
ing on when it occurs in relation to the sync or video component. If the
dropout pulse is derived between sync pulses, the dropout pulse is fed
directly through the dropout processor to the post-emphasis module. The
dropout pulse is applied to the DO clamp in the sync- channel video cir-
cuit, and, as a result, sync video is clamped at gray level during the
dropout.
If a dropout takes place during sync, the dropout pulse fed to the
dropout processor module is coincident at an AND gate with a 17- micro-
second pulse derived from the inverted leading -edge gate pulse. When
this happens, the output of the AND gate triggers the 2H/3 multivibrator.
The multivibrator in turn produces a pulse lasting two thirds of a line,
or 42 microseconds. This pulse added to the dropout pulse becomes the
stretched sync dropout pulse, which is fed to the DO clamp circuit in the
sync -channel video section of the post-emphasis module. The stretched
dropout pulse eliminates horizontal sync or, during the vertical interval,
two equalizing pulses.
Besides being fed to the post- emphasis module, the sync dropout pulse,
stretched or otherwise, is also fed to the sample -pulse eliminator in the
ATC error detector module. The sync dropout pulse is one of the two in-
puts to an AND gate; the other is the 9H pulse. The sample -pulse elimi-
nator produces a pulse that inhibits the ATC sample -pulse generator if a
dropout is present during the 9H interval. Thus, whenever the sync drop-
7:i7 TELEVISION BROADCASTING
DO FM
i
I n put FET FET
TP3
(Tape FM) Q8 Q7
FM In
(From FM Equal 523)
-oJ!
'
(From FM
Ref 508)-
FM +
-26 V I I
-11 MHz
Ref Out Low -Pass Filter
EE
(To FM Filte 524)
+20 V-4-0
J_v-_
Dropout
DO Test
(Play
Control
- Inverter
042
Insert
BSMV
Q43, Q44
X19
N
Panel)
Coincidence Circuit
-
Track Red
Play 617)
Sync DO Pulse
TP1
DO Pulse
AGC Control
DC Amp
013
DO FM Out
TO DO Comp 620)
EF FB Amp AGC Del
Q6 Q4, Q5 X7
R49
FB Amp EF
QQ, Q3
Ql Q12
S1
mi/ Sensitivity
Set 1
1 TP2
Sim DO
o
- Gate Driver
Q10
Phase
Split
Q9
Clamp
R45
Balance
AND Gate EF
WI Q30
9H In
(From Sync Logic 533)
Sync DO
On Voltage
ITo Post -Emp 529)
Coincidence Circuit
X2
X3 Q9 -C
Al
Inverter AND Gate
Q3
BSMV
Gate
04
-ro-
Q1
DO Vert
out pulse and the 9H pulse are coincident, the ATC sample pulse is elimi-
nated. This allows the ATC circuits to continue with their previous com-
pensation.
The 2H/3 multivibrator also generates another pulse from which is
derived a dropout vertical pulse. This pulse, which is fed to the Linelock
module, replaces the vertical -sync pulse normally used in vertical timing in
the event it is lost due to a dropout. When Pixlock is achieved, the Pix-
lock /Linelock bus is at ground potential. This potential and the 2H/3-
generator pulse furnish the inputs to an AND gate. During the period that
coincidence exists between the AND gate inputs, the AND gate is enabled,
and the succeeding amplifier produces the dropout vertical pulse.
The dropout vertical pulse is applied simultaneously with the tape ver-
tical signal to the tape -vertical /reference coincidence detector in the Line-
lock module. Whenever the vertical -sync pulse that is used in deriving the
tape vertical signal is eliminated due to a dropout, the dropout vertical
pulse is present as a substitute. Thus, the servo systems are prevented from
falling out of lock during a dropout in the tape vertical sync.
The simulated -dropout circuit permits the operation of the dropout -
compensation system to be checked by inserting dropouts in the fm sig-
nal. The simulated dropouts are inserted just prior to the fm output stage
that feeds the line to the Dropout Compensator module. Depending on
which signal, fm or video, is selected for observation, the dropouts will
appear as gaps in the continuous fm wave or as lines in the video. The
simulated-dropout insertion circuitry is activated by pressing the DROPOUT
INSERT switch on the front panel of the dropout processor module. To
view the dropouts in the signal on the cro or picture monitor, the COMP
OFF /MONO ONLY /NORMAL switch on the front panel of the Dropout
Compensator module must be in the Mono Only position.
The dropouts are placed in the fm signal by a diode quad (M1) which
is gated on during five alternate lines once in every frame. The quad gat-
ing signals are derived from the output of a three -input AND gate. One
of these input signals is a steady -state control signal. The second is a 10H
pulse timed to occur about four milliseconds after the beginning of each
frame. The third is a 30-microsecond pulse timed at half the horizontal
rate to occur about five microseconds after the start of a line. Conse-
quently, once every frame for a 10H period, all three AND -gate inputs will
be coincident. During this period, five 30- microsecond pulses (one every
other line) will be passed by the AND gate. These pulses are then split
in phase to become the quad gating pulses. The simulated dropout passed
each time the quad is turned on causes a reduction in the level of the fm
signal being fed to the Dropout Compensator module. This reduction in
level is sufficient to activate the dropout- detection circuitry in the latter
module. When the demodulated signal is viewed on a picture monitor,
five lines about one quarter of the way down the raster will be visible in
the display.
THE TAPE -DROPOUT COMPENSATION SYSTEM 245
The dropouts are eliminated by resetting the flip -flop that produces the
steady -state input to the AND gate. This is done by tapping any of the
following switches: STOP, WIND, PLAY, STANDBY, RECORD, SETUP, CUE
ON /OFF, or AUDIO ON /OFF.
graphs.
The SENSITIVITY control adjusts the sensitivity of the dropout detector
circuit. To the detector, a dropout is a reduction in the rf amplitude of the
fm carrier. As the SENSITIVITY control is turned clockwise, the sensitivity
decreases so that a greater drop in the rf amplitude is required to cause
the video to be replaced. As the control is turned counterclockwise, the
smaller a drop in the rf amplitude must be to cause the video to be re-
placed. The useful range of the SENSITIVITY control is the upper (clock-
wise) two- thirds of its range. This is 10 to 20 dB below the normal rf in-
put level of 0.4 volt. As the SENSITIVITY control approaches the maximum
counterclockwise position, the detector sensitivity becomes so great that it
causes excessive replacement of the video.
When determining the sensitivity setting, the operator must consider
the condition of the tape and the condition of the recorder. Amplitude
variations caused by the recorder, such as those resulting from slope detec-
tion or tracking variations, could be detected as dropouts if the sensitivity
is too great. The ideal sensitivity setting is one which compensates for all
dropouts, large and small, up to the point where the replacement video
causes an apparent change in the overall picture.
In general, if the rf signal level is consistent, operation at a higher
sensitivity is permissible. If the level is not consistent, it is preferable to
operate at a lower sensitivity. Under noisy signal conditions, the DOC
will, of course, supply more replacement video. If the sensitivity is too
great, the difference between the replacement video and the direct video
could be perceptible in the overall picture.
The mode selector switch has three positions that correspond to three
modes of operation: Normal (normal full compensation of NTSC color
signals or black and white) , Mono Only (monochrome only) , and No
Comp (no compensation) These functions are explained in the follow-
.
ing paragraphs.
In the Normal position, all functions of the dropout compensator are
operating. This is the position that should normally be used for 525 -line
or 625 -line NTSC color or for monochrome operation. The dropout corn-
246 TELEVISION BROADCASTING
pensator will replace both the luminance and the color subcarrier with the
stored video, in the event the reproduced rf level drops below that set by
the SENSITIVITY control. If the rf level is normal, the video is routed di-
rectly through the video switch to the post amplifier and the video output.
In the Mono Only position, the DOC operates as in the Normal position,
except that the color subcarrier is removed from the replacement video.
The direct video is not affected. That is, as long as there are no dropouts
and the DOC is supplying direct video, the chrominance signal and color
burst are unaffected. The color signal is eliminated only in the delayed
video inserted in the event of a dropout. The Mono Only position would
normally be used with color systems other than the NTSC color system.
In the No Comp position, direct video is always supplied from the
DOC. The detector circuit is locked in the no-dropout state, and the
video is routed through the direct -video side of the video switch and
amplified to unity gain in the post amplifier.
Operation of the Color Dropout Compensator is based on the concept
that lost information may be replaced by information from the preceding
scan line with little or no noticeable degradation of the picture. To do
this, the dropout compensator always has stored in its delay system the
previous scan line of video information. By continuous sampling of the rf
level of the reproduced fm signal, it may be determined if the video has
a significant loss of information, or if it is acceptable.
If the rf amplitude of the reproduced video signal is acceptable, as set
by the SENSITIVITY control, the signal is simply routed through the DOC
and back to the recorder output circuitry. At the same time, the acceptable
line of video information is sent through the delay system to be ready for
use if the following line has a significant dropout. If the rf level drops
below the level set by the SENSITIVITY control, the DOC switches from the
current video information to the previous line of video information, which
is stored in the delay system.
The direct uncompensated video is fed into one side of the video switch
circuit, and the delayed video is fed into the other side. The detector con-
trols which video is the output from the DOC. The video output is con-
tinuously fed into the delay system and stored for one line.
Fig. 8 -3 illustrates the layout of the DOC circuit boards and the signal
flow between them. All incoming signals, except the VI bus which con-
trols K1, come into the detector and video switch board. These input
signals are the uncompensated video, the fm input, and the DOC- enable
signal from the play bus. The detector monitors the fm for dropouts and
controls the video switch as previously described. For the detector to op-
erate, a ground enable signal must be provided on pin 15 by the play bus.
Also, as previously described, the detector provides a sync clamp pulse on
pin 29 when a dropout is detected. The video switch, controlled by the
detector, selects which video, the direct video from pin 18 or the delayed
video, will be the output.
THE TAPE -DROPOUT COMPENSATION SYSTEM 247
a
F
ó
>
co
á
o
Ti
o' É
L
ó
> o
>
G
E
o >
o
á
or
U
G E
o
m
U
E r
c` á
Courtesy RCA
Fig. 8 -3. Block diagram of color dropout compensator.
248 TELEVISION BROADCASTING
The output from the video switch goes to two places, the post amplifier
board and the luminance delay modulator board. The post amplifier am-
plifies the video output signal to provide an overall gain of unity through
the DOC. The video signal to the luminance delay modulator board goes
to relay K1. When the system is operating under the 525 -line NTSC sys-
tem, K1 is de- energized and the video signal is passed straight through.
When the system is operating under the 625 -line NTSC system, K1 is
energized, and the video signal passes through a 0.5- microsecond delay
line. From the relay and delay line, the signal goes to two places, the lumi-
nance delay modulator and the chroma delay board.
On the luminance modulator board, the video information is modu-
lated onto a 10 -MHz carrier. From the luminance delay modulator board,
the modulated signal goes to the luminance delay demodulator board,
where it is delayed one scan line of time and then demodulated. The de-
layed luminance then goes to the mixer circuit on the chroma delay board
to be mixed with the delayed color information.
On the chroma delay board, the video signal from K1 on the luminance
delay modulator board is filtered so that only the chrominance informa-
tion remains. The resulting signal is phase inverted and then delayed one
RF In
(From Demod Input Gate)
r
Dropout Sensor
R58 Dropout
Detector
R60
Level T1
Emitter
Limiter
Q12
Fol lower
Q15
lirDetecto 018
Low Pass
Filter
L1,L3,C20,C3
Emitter R42
Lim'ter Emitter
Follower Amplifier
Q14 Fol lower
Q16
013 Q19
P43
o Play
Emitter
0
Follower
Q10
oy
X6 Record
Tunnel
Diode
K2
Switching Signal Out Amplifier
10
To Demodulator Q11
scan line of time. The chrominance information is then mixed with the
luminance information, and the complete video signal, delayed one line,
is fed from the chroma delay board to the delayed side of the video switch
on the detector and video switch board.
The reason the luminance and color are delayed separately is that the
color signal of each succeeding scan line is shifted in phase by 180 °.
Therefore, the delayed video must have its color subcarrier not only de-
layed one line, but also phase inverted, as previously explained.
stage output amplifier; see Figs. 8 -4 and 8 -5. (One of the sections is the
calibration -pulse generator, which is not discussed here.)
There is a relay- selected rf signal at pin 6 of the demodulator input
gate, and this signal is applied to an input of the dropout sensor. This rf
signal (approximately I/2 volt peak to peak) enters the dropout sensor
(module 6) at pin 22 and is routed through dropout- detect level control
R58 to the base of limiter Q12. The output of Q12 is coupled by emitter
follower Q13 to the base of second -stage limiter Q14. The amount of limit-
ing is a function of the signal level applied to Q12, which in turn is a
function of the setting of potentiometer R58.
The output of Q14 is coupled by emitter follower Q15 to full -wave
rectifier (or detector) Q17 -Q18 through transformer T1, which provides
signals of opposite phase at the bases of Q17 and Q18. Because both are
pnp transistors, when one turns on, the other turns off, resulting in full -
wave rectification at their common collectors. As long as the level of the
incoming rf equals or exceeds the limiting level, the output of Q17 and
Q18 is a straight line. If the level of the incoming rf falls below the
limiting level, both transistors turn off, and their collectors swing toward
ground (zero volts) .
The limited and full- wave -rectified signal at the output of Q17 -Q18 is
routed through a low -pass filter to the base of emitter follower Q19. The
emitter follower remains cut off as long as the positive voltage of the sig-
nal from the collectors of Q17 and Q18 remains at the limiting level.
When, however, this voltage falls below the limiting level, Q19 turns on,
and its emitter voltage swings negative. The result is a negative pulse the
duration of which coincides with the duration of the dropout period
(Fig. 8 -6)
.
During the play mode, the output of Q19 is routed through de- energized
relay K2 and coupled by emitter follower Q10 to the base of switch Q11.
The base circuit of Q10 includes tunnel diode X6, which acts to greatly
accelerate the rise times of the negative dropout pulses applied to the
250 TELEVISION BROADCASTING
R50 R49
47 6800
41
C16 47 C6 15
I20V T 20V
20 20 = 9 =
22
When the rf level equals or exceeds the limiting level, Q16 is satu-
rated, and the effective emitter resistor for limiter Q14 is R42 (150
ohms). Thus, under this condition the gain of Q14 is approximately 10
(i.e., R56, 1500 ohms, divided by R42, 150 ohms). When the rf level
falls below the limiting level (i.e., the dropout condition) , Q11 turns on
and biases Q16 off. Under this condition, the effective emitter resistor for
Q14 includes R42 plus R43 (240 ohms), and the gain is 1500/390, or
approximately 3.8.
When the gain of Q14 falls from 10 to 3.8, the signal level at the base
of Q15 falls, removing any possible ambiguity in the indication of a
dropout that is marginal. Thus the presence of amplitude modulation
cannot cause a marginal dropout to appear to be a cyclic alternation be-
tween a normal and a dropout condition. When any dropout occurs (e.g.,
rf level down 20 dB from normal), Q11 will remain on, Q16 will remain
off, and the gain of Q14 will remain at approximately 3.8 until the in-
TI
Dropout Detector
018
202048
+12V
R48f 47
18
1
C29
3pF500V
L3
39 pH
C19
12vF500V
LI
27 pH
E
Pulse Stretcher
C21
20 =
OI
1OV
R33
12 V
18
5600
R34
Q19
2N2048
43
20 = 20
- 500V
20 =
19
-12V
18
fill R32 75
Q10
2N2048
2N1308
R31 1000
X6
K2
114371 :
R26 5600
936 820
16 - 9
-I2V
47
R52}47 C14
20V
19
-1 V
1171151141131 EI El } 11 1E1
Nd Used Gnd Cod Switching From EE Tape
Signal Out Cal -Pulse Control
Section of In
Board 6
of dropout sensor.
252 TELEVISION BROADCASTING
coming rf level is down only 11 dB from normal. At this point Q11 will
turn off, Q16 will turn on, and the gain of Q14 will rise to the normal
level of 10.
Dropout
Condition
Output of 015
RF In (Pin 22( Bases of 017 -018)
Base of 019
Filtered Output
Limiting Level (0 V)
of Q17 -Q18
019 Off
Emitter of 019
Q19 On
011 On (0V)
Collector of 011
EXERCISES
Chapter 6 traced the need for special compensation systems for mono-
chrome geometric error, and the additional types of correction necessary
for the color information. These corrections act primarily on a head-to-
head basis, or on the entire group of 16 to 17 lines occurring within one
head pass across the tape.
There are additional color errors which are not "cleaned up" by the
previously discussed systems. These are:
1. Hue (phase) shifts within each band of 16 to 17 lines of a head
scan.
2. Saturation (amplitude) variations within each band of 16 to 17
lines of a head scan.
The advanced color -error compensation systems required to correct these
errors are the Velocity Compensator, used in conjunction with the Colortec
and Amtec (Ampex) , and the Cavec (Color Amplitude and Velocity
Error Compensation), used in conjunction with the monochrome and
Color ATC systems (RCA) .
The line -by -line correction is made possible by the basic assumption
that errors found in one revolution of the headwheel will also be present
in the next revolution of the headwheel. During the head revolution, each
of the four heads scans 16 lines; thus, to store information for one com-
plete revolution, 64 memory units are required. Separate memory units are
required for the velocity and chroma-amplitude corrections. Thus, the
memory units in any one Cavec total 128. The memory units are really
capacitors which are charged in approximately 100 milliseconds, or 24
revolutions of the headwheel. It should be recognized that instantaneous
changes in either hue or saturation will not be corrected to some absolute
value but will be corrected to the average value of the previous 24 revo-
lutions of the headwheel.
The CAC corrects for frequency -response variations in the video sig-
nal, which produce saturation errors in the color picture. The VEC cor-
rects for inaccuracies in head-to -tape velocity, which cause undesired hue
changes.
The circuits of the Cavec are similar to those used in computers. Some
are of the analog- computer type, and others are of the digital- computer
type. As shown in Fig. 9 -1, the analog circuits are divided into two
groups, one for CAC and the other for VEC. The digital circuits are a
single group shared by both VEC and CAC. The output signals which
correct the errors are produced by the analog circuits. The digital circuits
provide the analog circuits with timing and control pulses required for
line -by -line correction.
The input to the CAC (Fig. 9 -2) consists of burst separated from the
demodulated video. In the analog circuits, the burst amplitude is com-
pared to a reference level, and the resulting error voltage is stored in a
2 TELEVISION BROADCASTING
o
ii
E
G
o
E
ó
dlñ°5
CJV QCJO
iiit-
E
o
- x
Tape FM In-. FM
Equalizer
Module
FM
Demodulator Tape Video
±V Equalizer
Control Voltages
Filtered
Cavec Burst
Unit Filter
(CAC)
Burst Separator
The VEC portion improves the monochrome ATC system of the ma-
chine by causing the MATC variable delay line to correct timing errors
continuously instead of in steps. In machines without Cavec, the MATC
error signal is applied to the variable delay line at the beginning of each
line and remains constant throughout the line. The error waveform there-
fore appears as a series of steps, as shown in Fig. 9 -3. This type of signal
restores video timing to normal at the beginning of the line, but it allows
errors to accumulate during the line. As a result, the chroma shifts be-
tween correction steps, causing hue banding or, in the case of PAL opera-
tion, desaturation.
To correct this deficiency, the VEC adds ramps ( Fig. 9 -3) to the MATC
error signal. The resulting waveform closely approximates the true, con-
tinuously variable velocity -error signal. To produce ramps of the correct
slope, the VEC analog circuits measure the differences between the suc-
cessive step voltages and store the difference voltages in a capacitor mem-
ory similar to the one used by the CAC. The memory contains 6.4 capaci-
tors, one for each of the 16 lines in each of the four head scanning periods.
(Since some head scan periods contain seventeen lines, the error stored in
the sixteenth capacitor is used for both the sixteenth and seventeenth
lines.) The time constant of the capacitor charging circuit causes the
error voltage to stabilize after 24 headwheel revolutions. Each stored volt-
age, therefore, is an average for 24 revolutions. At the beginning of each
258 TELEVISION BROADCASTING
Actual
A
Timing Error
MATC Error
VEC Ramp
VEC Output
Expanded View
(Sum of
D
Waveforms
B and Cl
_17-Line 16 -Line
Head Scan Head Scan
The digital system occupies the first 12 board positions. This system
actually contains 11 boards, but only one of the first two positions is used,
position 2 in the TR -70 and position 1 in all other machines. Only five
different types of digital boards are used. The principal components on
these boards are integrated- circuit packages of the dual in -line type. The
packages contain logic circuits, such as gates and inverters, or timing cir-
cuits, such as one -shots or flip-flops. All logic elements are of the diode -
transistor logic (DTL) family.
Positions 15 through 18 are occupied by four identical boards which
constitute the VEC memory. These boards contain storage capacitors,
transformers, transistors, and semiconductor switches. A total of 64 stor-
age capacitors is provided, 16 on each board. Each capacitor stores infor-
mation pertaining to a particular line in a particular head scanning inter-
val (for example, line 1 of the head -1 interval).
Positions 13, 14, and 19 contain the VEC analog boards. The compo-
nents on these boards consist mostly of semiconductor switches and inte-
grated- circuit operational amplifiers. The operational amplifiers are basi-
ADVANCED COLOR -ERROR CORRECTION SYSTEMS 259
The velocity -error corrector (VEC) contains two main groups of cir-
cuits, as shown in Fig. 9 -4. The first group, called the analog system (Fig.
9 -4A) , consists mostly of operational amplifiers and semiconductor
switches. The second group (Fig. 9 -4B) , called the digital system, contains
pulse generators, counters, gates, and other forms of digital logic circuits.
The functions of the analog system are to generate a series of ramp sig-
nals (straight -line segments) , one for each horizontal line in each of the
four video -head scanning periods, and to add the ramps to the MATC
error signals. The MATC error signal consists of a series of steps which
change level at the beginning of each line. The amplitude of each step
constitutes a measurement of the velocity error existing at the beginning
of the corresponding line. The true velocity error, however, is a smoothly
changing signal, rather than a series of steps. Addition of the ramps
smooths out the steps to provide a better approximation to this signal.
260 TELEVISION BROADCASTING
E
U
L J i
2 C
1- N
t. 2C
11
an
E E
E Q
N
The clamped error signals are amplified and then fed to a semiconductor
switch called the write switch. Once per line, for a short interval before the
clamping operation, a pulse from the digital system closes the write switch,
thereby applying the signals to a bus (input /output, or i /o, bus) which
goes to a bank of memory capacitors. As explained later, storage of the
error signal differences in the memory bank provides information which
Sum of
MATC
And Line 10 Line 11 Line 12 Line 13
I
I
CATO
Error
Signals E10-V11-V10
i
O 1
V10 E12
V13
VII
I
1
IV12 (
CD
Clamp
Pulse
0
Ell
O
1
E10
Clamp
Output
I_ (((%%%rrrJJJ
E12
Read
Pulse
0
O
Ramp
E10 E11
E12
O
VEC
Output
(Ramp
& MATC)
Solid
Line
or -E10, the value of E10 must be predicted at the start of the ramp, from
the information available at that time. This information consists of the
voltage previously stored in capacitor C10, since this capacitor will not be
written into again until the next line.
During the read -out period, the voltage from the addressed memory
capacitor is applied through the i/o bus to a driver amplifier (Fig. 9 -4).
The driver output is connected to a semiconductor switch called the sample
switch. For a portion of the readout time, the sample switch closes and
transfers the driver output voltage to a capacitor which serves as a short -
term, or temporary, memory. The sample switch and capacitor are referred
to as the sample -and-hold circuit, because the capacitor retains the sample
voltage after the sample switch opens.
The voltage from the sample-and -hold circuit is applied to an integrator
circuit, which charges a capacitor at a constant rate proportional to the
sample voltage, thus generating the required ramp. The MATC signal is
added to the ramp in a summing amplifier to provide the VEC output
signal, which is fed to the MATC circuits of the recorder.
26 TELEVISION BROADCASTING
The memory -control pulses consist of two types, called X and Y signals.
For the VEC system, there are sixteen X signals (VX1 through VX16) and
four Y signals (VY 1 through VY4) One X signal is provided for each
.
-i 15 us
OS1
Ji
1 0S2 1
r
Tape Hor iz Input
If Required
15 3
o
v
O
15µs
OS1
"1" Output
0S2
"1" Output
(B) Waveforms.
Fig. 9 -6. Use of one -shots in cascade.
at the negative -going (leading) edge of each horizontal pulse, the first
one -shot is triggered, and its 1 output produces a positive -going pulse.
At the end of the 15- microsecond pulse period of this circuit, the 1 output
falls to the low level, causing the second one -shot to be triggered. The 1
output of this circuit then produces a positive -going pulse lasting for three
microseconds, as desired.
The preceding method is extended, in the actual equipment, by connect-
ing more than two one-shots in cascade. For example, the block diagram of
Fig. 9 -7 and the associated timing diagram of Fig. 9 -8 show in simplified
form how the clock, clamp, read X, write, and write X pulses are generated.
The tape- horizontal pulse is fed only to the first circuit in the chain,
which is called the clock one -shot. When triggered by the negative -going
266 TELEVISION BROADCASTING
edge of tape horizontal, this circuit produces a positive -going 0.5- micro-
second pulse at its 1 output and a negative -going pulse of the same width
at its 0 output. The purpose of this circuit is to produce narrower pulses
than the tape -horizontal pulses; the narrower pulses serve as more con-
venient references for the other one -shot circuits.
Both the 1 and 0 outputs of the clock circuit are used as reference. The
1 output produces a delay equal to the pulse width, or 0.5 microsecond,
because its trailing edge is negative going. The 0 output, however, pro-
duces no delay because its leading edge is negative going.
Clock
OS
-ycT
OS
Read
X
- Clock
Read X
0.5 4.3
OS
Ramp
I- Ramp Reset
rCT
Reset
-
1.9
Write
OS OS 1 Clamp
t
OS 35 10
Write
-SCT Delay
8.5 os OS
1- +WriteX
Write X Write
Delay X
0
18 15
Among other circuits (not shown in Fig. 9 -7) , the positive -going clock
pulse is applied to the read one -shot. This circuit produces a 4.3- micro-
second pulse, which is used in the read portion of the memory cycle.
The 0 output of the clock one-shot is fed to the write -delay one-shot,
which produces opposing pulses 8.5 microseconds wide at its two outputs.
As in the clock circuit, both outputs of the write -delay one -shot are used
as timing references. The 1 output produces a delay of 8.5 microseconds,
but the 0 output produces no delay.
The 1 output of the write -delay one -shot triggers the write one-shot,
which produces a positive pulse 35 microseconds wide at its 1 output. This
pulse operates the write switch of the analog system. The trailing edge of
ADVANCED COLOR -ERROR CORRECTION SYSTEMS 267
the write pulse triggers the clamp one -shot, which has a pulse width of 10
microseconds. The 1 output of the clamp one -shot operates the clamp
switch of the analog system. The clamp switch, therefore, closes as soon
as the write switch opens, and it remains closed for 10 microseconds.
The 0 output of the write -delay one -shot is used as a timing reference
for the write-X delay one -shot, which has a period of 18 microseconds.
The 1 output of the write -X delay circuit triggers the write -X one -shot,
Time (Microseconds)
0 10 20 30 40 50 60
Tape Horiz
Clock 0.5us
Write 35 us
Clamp 10us
-t
Write X I- 15us -.
Fig. 9 -8. Simplified timing diagram of horizontal -rate timing pulses.
J T
Le,
_ AIL
JO
Á Ñ
Ñ
>
cn
N
Ir
>
O.
LT-
each consisting of one X and one Y signal. Each pair corresponds to one
of the 64 capacitors in the memory bank. A particular capacitor is addressed
when its corresponding X and Y signals occur simultaneously. The X signal
closes a switch that connects one side of the capacitor to the i/o bus, and
the Y signal closes another switch that connects the other side of the ca-
pacitor to ground.
Information for generating the X signals is derived from a line counter
and a decoder. The counter provides information as to which line is being
scanned in any of the four head scanning intervals. The X decoder combines
this line -count information with the read and write pulses to determine
which capacitors should be connected to the i/o bus during the read and
write intervals. (As previously mentioned, in any one line, different capaci-
tors are addressed during the read and write periods. For example, in line
11, C11 is addressed during the read period, but C10 is addressed during
the write period.)
One Y- signal interval is provided for each of the four head scanning
intervals. Information for generating the Y signals is obtained by using a
decoder to extract from the 4 x 2 and 2 x 1 switcher outputs of the re-
corder information that indicates which of the four heads is scanning the
tape. This information is combined with the read and write pulses to
form the four Y- interval signals.
To permit addressing the memory capacitors in the desired sequence, a
switching matrix is used as shown (simplified) in Fig. 9 -10. There are 16
control buses, VX1 through VX16, from the X decoder, and four control
lines, VY1 through VY4, from the Y decoder. The prefix "V" is used to
distinguish the control lines of the VEC memory from the corresponding
lines of the CAC memory. The bar over the VY symbols indicates that the
switches controlled by these buses are closed when the signal voltages are
at the lower of the two possible levels.
An individual normally open switch is connected between one side of
each of the 64 capacitors and the i/o bus. Each VX bus simultaneously
controls four of these switches. Each VY bus controls only one switch, but
each of these switches is connected between ground and 16 capacitors, one
from each of the 16 VX groups.
Normally the VX buses are low and the VY buses are high, so that all
switches in the memory are open. When a given capacitor is addressed,
however, the corresponding VX bus goes high, thereby closing four
switches and connecting four capacitors to the i/o bus. Simultaneously,
the corresponding VY bus goes low, thereby closing the Y switch corre-
sponding to the head scanning interval, which grounds 16 capacitors.
Only one of the 16 capacitors, however, belongs to the group of four con-
nected to the i/o bus by the VX switches and, therefore, has a complete
charge path. All other capacitors are inactive, because one side of the
charge path is open. For illustration, the charge path of the capacitor cor-
responding to VX2 and VY4 is shown in heavy lines in Fig. 9 -10A.
27O TELEVISION BROADCASTING
110 Bus 4i
ar
VX1
VX2 -1--
VX16
X Drive
scanning line, and the second pulse connects the same capacitor to the i/o
bus during the write time of the next scanning line. This is done to permit
predicting the slope of the ramp required for connecting the two MATC
steps.
ADVANCED COLOR-ERROR CORRECTION SYSTEMS 271
Like the VEC, the CAC consists of an analog system and a digital system.
The analog system compares the amplitude of the burst in the tape video
signal with a threshold level, generates an error voltage proportional to
the difference, and stores the error voltage in a capacitor memory. At the
beginning of each line, the voltage on the corresponding memory capacitor
is read out and used to generate two push -pull control voltages. These
voltages are fed to the fm equalizer of the tape machine, where they auto-
matically vary the equalization to keep the burst level constant.
272 TELEVISION BROADCASTING
Read' l Write
VX1 11
VX2 Il II
Il II
II 1
1 I I I
Il II
I II
ll II
II II
II I I
II I
II LL
II lI
II II
II II
VX16
I II III
Head Scan I
For 17 Line Pass Only
Vl'1
11IIIIII11111111III
Head Scan 2
VY2
1111111111111111111
_
VY3
11111111'
_
VY4 Sti
Illllf
Although the CAC error voltages are based entirely on the burst ampli-
tude, the burst actually constitutes a sample of the chroma signal. If there is
no differential gain in the system, the changes in equalization derived from
measuring the burst amplitude alone will be correct for the entire chroma
signal.
As in the VEC system, functions such as reading out of or writing into
the capacitor memory are controlled by semiconductor switches. These
switches are operated at the required times by pulses from the digital
system. Many of the digital pulses are identical to those used in the VEC
and are produced by the same circuits.
-bc+i->
O _
_
o _
wá U
o
T
J
L
oce
U
._.
a
N
o o
706
LAI
Cr
y
v0
a U
Q
N
O _
m > >
E
m
vó
L)
in x
to
ai .
L p o
7
CO
7
Qm
Ñ O
r
U
L L
read out by applying it to the i/o bus. During the read interval, a switch
transfers the voltage from the i/o bus to a temporary- memory capacitor
in the sample- and -hold amplifier. The output of the sample-and -hold
amplifier is fed to two drivers which develop push -pull control voltages
(± V) proportional to the sampled error. These two signals constitute the
CAC control output of the Cavec unit.
Clock
red
Filteurst 11111
B
Half Wave
Detected
Burst
1II1 .11111 IIIII
Low-Pass
Filtered
Half Wave
Burst
Peak
Detected
Burst
Sample
Write
Clamp C
Peak
Detected
Burst
V
tV
Control
FM Input to Equalizer
in the picture because a seventeenth line does not occur periodically during
successive scanning periods of the same video head.
This circuit is inserted in the path of the delayed signal in place of the
potentiometer. The ± V signals from the CAC output of the Cavec are
then used to control the attenuation. When the manual /auto relay ( Fig.
9 -17) is in the auto condition (energized) , the ± V signals are derived
from the error signals stored in the CAC memory. The equalization is thus
controlled automatically on a line -by -line basis. Under these conditions,
the manual equalization control on the equalizer module is disconnected,
and the only way that the degree of equalization can be changed manually
is by adjusting the BURST RATIO control on the front panel of the Cavec
unit. When, however, the relay is in the manual condition (de- energized) ,
the ± V signals are derived from a dc voltage which depends on the setting
of the manual equalization potentiometer on the fm equalizer. The equali-
zation can then be controlled manually by adjusting this potentiometer.
Fig. 9 -18 is a schematic diagram of the fm attenuator board. The circuit
consists of diode quad Ml, transistors Q1 and Q2, and two regulated sup-
plies consisting of R4, X1, and Cl and R7, X2, and C2.
Direct
Signal Path
Delay Line Equalizer Difference Equalized
FM Input-N. Driver Delay Line Amplifier FM Output
Reflected
Signal Path
Diode Quad
FM In (Attenuator) FM Out
+V Signal
Cavec Curren
(CAC) Source
-V Signal:
FM Attenuator Board
f = FM Crossover
° Frequency
2f0
0
FM Frequency (Input)
An output voltage of -3.9 volts is provided by R7, X2 and C2; R4, X1,
and Cl provide +3.9 volts. Transistor Q2 (npn) diode quad Ml, and ,
transistor Q1 (pnp) are connected in series between the -3.9- and +3.9-
volt supplies. The delayed fm signal is fed into junction A of Ml, and the
output is taken from junction B. The +V signal from the Cavec module is
applied to the base of Q1, and the -V signal is applied to the base of Q2.
The networks containing R1, C3, R3, and C4 filter the ±V signals. The +V
r
Direct
Equalizer I 2 3 4 5
FM Attenuator
Manual Board
EqControl
(On Eq Module -3.9 V
Front Panel) Current
Source
FM Equalizer
+V Control -V Control
Manual
o .
Push-Pull
Driver
Circuits
CAC Error 1 o f:
Manual IOff l
Auto
and -V signals control the current through the path consisting of Q1,
junctions C and D of M1, and Q2. The current decreases as the +V signal
assumes more positive values ( and the -V signal assumes more negative
values) because the transistors approach cutoff. For fixed values of +V
and -V, the current is constant.
Since the four diodes are forward biased, the fm signal applied to junc-
tion A can flow through the quad to the output at junction B. The im-
pedance between junctions A and B, however, depends on the amount of
forward bias current. This, in turn, depends on the current supplied by QI
and Q2. The quad, therefore, attenuates the fm input by an amount depend-
ADVANCED COLOR -ERROR CORRECTION SYSTEMS 281
ing on the ±V signals. The larger the +V signal, the greater is the
attenuation of the delayed -path fm signal and the greater is the chroma
content of the video. Conversely, the smaller the +V signal, the smaller are
the attenuation and the chroma content of the video. Use of the two bal-
anced push -pull drive signals reduces transient disturbances and distortion
of the fm signal.
+12V
+3.9V
C1y
+V Control
From CAC
I
FM Input FM Output 1
R2 R8
-V Control
From CAC
3 C4
For Manual
3.9 V Equalization
Control I n
FM Equalizer
-12V
The -3.9 -volt zener regulator circuit (R7, X2, and C2) , in addition to
providing bias to Q2, provides the input voltage to the manual equalization
potentiometer on the fm equalizer module. The arm of this potentiometer
is connected through a resistor divider to the manual /auto relay in the
Cavec unit. When the relay is de- energized, the attenuated potentiometer
voltage is fed to the input of the first push -pull driver stage. Consequently,
since the ±V signals depend on the potentiometer voltage, the equalization
can then be varied manually by adjusting the potentiometer. When, how-
ever, the relay is energized, the potentiometer voltage is disconnected, and
the CAC error voltage from the sample- and -hold capacitor is fed to the first
push-pull driver. Under these conditions, the potentiometer has no effect,
and the equalization is controlled automatically by the CAC signals from
the capacitor memory.
282 TELEVISION BROADCASTING
EXERCISES
Motion -picture film can be examined on a frame -by -frame basis at zero
speed for the purpose of cutting and splicing for editing. Video tape, how-
ever, produces no output unless it is moving, and it must be moving near
the recorded speed for a usable picture. Thus video -tape editing is essen-
tially a dynamic function rather than a static one, and it requires special
considerations.
It is natural that the first approach to video -tape editing or splicing was
a mechanical operation which involved making the magnetic tracks and
editing marks visible ( see Fig. 2 -10 in Chapter 2 for an example) so that
tapes could be cut at the proper place to avoid picture disturbances. Me-
chanical splicing is still used in limited applications, and we will consider
this technique first. We will then examine the more recent systems of elec-
tronic editing, and, finally, we will study automatic (computerized) editing
programmers.
First, it should be known that early tape systems (monochrome) did not
operate synchronously with local station sync, and 60 edit pulses per second
(one pulse per field) seemed most appropriate. However, when synchron-
ous operation became a practical reality, a specific field was used for refer-
ence so that splicing always occurred on the same field to avoid a momentary
half -line 'bobble." Therefore, the edit -pulse rate became 30 pulses per
second, and these pulses are termed frame edit pulses. Later, with the ad-
vent of color, the edit-pulse rate became 15 pulses per second in many
systems so that the four -field cycle of the NTSC color signal could be
accommodated.
(A) Stop recorder at desired location (B) Develop tape with solution of
and cut tape. iron oxide to locate nearest edit or
frame pulse.
(C) Align edge of video track on (D) Bring cutter into position and
built -in reticle to cut tape halfway be- make cut.
tween adjacent video tracks.
(E) Drape tape from supply reel over (F) Align tape track prior to
tape support post and position under making cut.
left hold -down door.
(G) Shear tape to be joined to tape (H) Dispense splicing tape and
already in splicer. prepare to put it in place.
these elements or a combination of them can cause splice failure and exces-
sive tape and head wear.
The adhesive used in a splicing tape is pressure sensitive. This special
adhesive, to assure a perfect bond, requires adequate pressure during appli-
cation. Occasionally, an operator does not apply sufficient pressure to the
entire surface of the splicing tape, causing an improper bond between the
video tape and the splicing tape. After the splicing tape is applied to the
(I) Insert splicing tape beneath (I) Press both ends of the tape firmly
the lifted ends of the two tapes to provide a perfect butt joint.
to be joined.
(K) Trim away excess splicing tape (L) Examine the completed splice.
with trimming tools.
Courtesy RCA
Fig. 10 -1. Steps in mechanical splicing of video tape. -cont'd.
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 287
video tape, pressure may be applied across the entire surface of the splicing
tape with a clean blunt instrument or clean rubber roller.
Proper alignment and centering of the splicing tape directly over and
in line with the splice cut must be assured for maximum strength. If the
splicing tape is not centered over the spliced sections, a lopsided splice
will result, with the narrow side being weak and susceptible to failure. If
any of the splicing -tape edges are not tightly sealed to the video -tape
backing, contamination will tend to collect at the areas of exposed adhesive.
Contamination deposits may cause tape damage and head wear in a manner
similar to the action of air bubbles.
Contamination of video-tape splices is caused by a variety of sources such
as solvents used for cleaning the tape, fingerprints, debris from transports
and splicing fixtures, and airborne dust.
Weakening and subsequent failure of splices can occur if the solvent
used to clean the video tape ( such as that used to remove splicing powder)
is permitted to seep onto the adhesive of the splicing tape. The finite
separation which exists between the sections of video tape being spliced can
allow the solvent to deteriorate the adhesive. Most commonly used solvents
(alcohol, Freon TF') are low -viscosity solutions which can easily penetrate
the small splice separation. To avoid solvent damage to the adhesive, be
certain that the tape is completely dry before applying the splicing tape.
Fingerprints, debris, and wear products can weaken the bond between
the splicing tape and the video -tape backing. A fingerprint is a deposit of
body oil which greatly reduces the bonding strength of the pressure- sensi-
tive adhesive. Fingerprints on either the video tape or the adhesive surface
of the splicing tape can be eliminated by proper handling or by wearing
lint -free gloves.
Ideally, the video -tape backing should be cleaned with a cotton swab
moistened with Freon TF to remove fingerprints and contamination. After
cleaning, the splice area should not be touched with exposed fingers. When
the solvent has evaporated and the tape is completely dry, the video -tape
sections can be aligned and the splicing tape applied and secured.
Cleanliness of the splicing fixture and the immediate work area must
be conscientiously observed during splicing operations. The knife portion
of the splicer and the splicing -tape application plate must be kept meticu-
lously clean to avoid trapping contamination within the splice. In addition,
the loose end of the splicing tape which protrudes from the dispenser is
subject to airborne contaminants and loose debris that can make it unsuit-
able for use.
In handling the splicing tape, avoid touching the adhesive surface which
will contact the video -tape backing, protect the splicing tape from con-
tamination, and do not distort or stretch the splicing tape during
application.
'DuPont Trademark
288 TELEVISION BROADCASTING
Approx 9 Inches
1/1111111111/1/1
Line S - 18 TV Frames
-; 1-8 ms
Video Record Heads
i' Video Erase Head Tape Travel -- A
'l
i 111111
Frame Intervals
Line S
18 Ti Frame
This line is further defined as existing in the guard band following a video
track that bears vertical- synchronizing pulses, as shown in Fig. 10 -2C.
In the operation of the Electronic Editor, the location of line S is deter-
mined after a random cue by means of a vertical -gating circuit, which
detects the first frame synchronizing pulse following the cue. At this
instant in time, the tape lies under the heads as shown in Fig. 10 -2D. The
times taken for line S to travel to the video erase head and the video record
heads may be defined as 8 milliseconds and 18 frames, respectively.
The 8- millisecond delay is obtained from a delay multivibrator; at the
end of this period, the erase head is energized by an electronic gating
circuit. Turn -on time is approximately 30 microseconds, and erasure starts
exactly on line S (Fig. 10 -3) The 18 -frame delay is computed by a binary
.
counter system. At the end of this period, the record current is gated on
electronically. Turn -on time in this case is 2 microseconds, and the new
recording starts immediately following line S, as shown in Fig. 10 -4.
Start of Erasure
If the action is repeated and the RECORD button is pressed, the scene con-
taining the production error is replaced with a new scene.
Obviously, in the insert mode it is necessary to make two splices per
operation and still maintain synchronism. The first (or ingoing) splice
was described previously. The second (or outgoing) splice is made in a
similar manner, except that at the end of the measured time intervals, the
erase and record currents are turned off instead of on. This is performed
with the same precision as that of the first splice.
Tape Travel -I. Line S
Splice
During the recording of the insert, the capstan -oscillator frequency is not
controlled; therefore, it may drift slightly if the inserted material is of
considerable length. This drift, in turn, will cause a variation in the wave-
length of the recorded control -track signal and result in an abrupt phase
shift in the reproduced control -track signal at the outgoing splice.
One solution to this problem would be to employ a phase- correction
system in the capstan -oscillator output circuit during recording. This, how-
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 293
ever, would be extremely complex. The adopted solution utilizes the fact
that in the insert mode a control -track signal already exists on the tape.
This information would normally be erased by the full-width video erase
head; however, a new head which has a separate section for the control -
track erasure has been developed (Fig. 10 -6). The control -track section is
disconnected when the insert mode is selected, and, although the machine
is placed in the record mode, reproduction of the control track continues.
Thus, it is possible to retain normal control of the capstan oscillator, and
any possibility of control -track phase shift is avoided.
33' (Exaggeratedl-
Control -Track
Erase Head
The video erase head is a new design which gives an improved erase
efficiency. It uses only a half turn in contact with the tape, and the elec-
trical gap width is 5 mils, the same as that of a guard band. The gap is
optically straight and inclined to the perpendicular at an angle of 33
minutes of arc. Thus, the erasure pattern is exactly parallel with the video
record pattern. The audio and cue tracks are not affected by the new head,
and control -track erasure is optional.
It will be seen that the functions of the Electronic Editor are of the logic
type and serve to provide precisely timed and electronically gated video
erase and record currents. A block diagram is given in Fig. 10 -7. The
various delay multivibrators, flip-flops, gates, and binary counters are all
294 TELEVISION BROADCASTING
classic circuit configurations that will not be discussed at this time. The
overall functions, however, may be traced as follows.
Following the initiation of the record mode, a delay multivibrator pro-
vides a 60-millisecond delay during which all normal record relays have
time to operate. Flip -flop 1 then operates and places AND gates 1 and 3
in a ready state. Pulses, derived from the Intersync unit, that mark the third
vertical-sync pulse in every frame interval are reshaped in a pulse former
and routed to AND gates 1 and 2. Gate 1, which is in a ready state, there-
To
Record
Relays
Record
1 Flip -Flop 1 AND Gate 1
Variable
One -Shot 1
Binary
Counters
AND Gate 3
8 ms
Frame
Pulses Variable
One- SOW 2 Flip lop 3 ---tow Erase
Switch
Flip-Flop 4 RF Switch ---rlo
570 us Record
Driver
To
Erase
1 Stop
Pros
fore operates at the first frame pulse and triggers variable delay one -shots
(multivibrators) 1 and 2. One-shot 1 serves to provide the necessary 8
milliseconds of delay before the video erase current is turned on. Erasure
turn -on is acomplished by flip-flop 3 and an electronic switch.
Because timing is referenced to the third pulse in the vertical-sync train,
and at this time the rotating video -head drum is positioned to place the
active head tip at the center of a video track, a 570- microsecond delay is
provided by one -shot 2. This delay allows the head to travel to the end of
the track, at which time the succeeding head on the drum periphery is
positioned at the beginning of the next track. From this timing reference,
18 television frames are counted off by the binary counter system, and at
the end of this period gate 3 operates. Video record current is then turned
on by the action of flip -flop 4 and a second electronic switch.
Timing for the outgoing splice is provided by the operation of gate 2,
gate 4, and variable delay one -shot 3. The operation of these stages, to-
gether with the reuse of the counter train, resets flip -flops 3 and 4 and
places the two electronic switches in the off condition.
An additional flip-flop (not shown in the block diagram) is operated
by the binary counters exactly 18 frames after the video record current is
turned off. This permits the audio track to be cleared of all extraneous video
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 295
information by the action of the audio erase head. (New audio information
has, of course, accompanied the new video information.) A time -delay
circuit allows all audio functions to revert to normal before the machine
is stopped. In this way, any possibility of transients on the audio track is
avoided.
The quality of the finished splice can be observed in Fig. 10 -8, which
shows the video, blanking, and synchronizing information in the video
waveform at the moment the electronic splice passes the video heads. The
splice appears the same as a change of picture caused by camera switching.
The rapidity with which the editing function is performed becomes
apparent when it is realized that the entire splicing operation is performed
while the recording medium is in motion at normal speed.
animated recording are made on the same machine, no tolerances are in-
volved, and the overlap can be eliminated by an adjustment.
The overlap for an ingoing splice is normally adjusted for 11/2 tracks
with the supply reel fully loaded and the tape running at normal speed.
The splice timing circuits then automatically change the overlap to 21/2
tracks for an outgoing splice. When the tape is running at half speed, the
overlap is automatically doubled. Thus, with the preceding adjustment,
the overlaps become three and five tracks.
Formation of Splices
An ingoing splice is formed by applying the video erase and record
currents to the respective heads at different times, instead of at the same
time as in normal recording. This is necessary because the master erase
head (in the RCA system) is approximately 7.5 inches from the headwheel,
and, if the two currents were applied simultaneously, the two signals would
overlap, without erasure, on the entire length of tape between the heads.
Similarly, formation of an outgoing splice requires that the video erase
and record currents be shut off at different times since, if they were shut
off at the same time, the section between the heads would be erased but
not rerecorded.
The timing of the erase and record functions during formation of a
splice is automatically determined by an electronic counter which, in effect,
counts the recorded frames on the tape as they pass the headwheel (Fig.
10 -10) The counter controls separate delay circuits for recording and
.
erasing. In turn, the delay circuits control transistor switches that either
allow or prevent application of the currents to the heads.
c
Q
~ T rT
zo
TO
O E ç
G-
-d ú
Uo -
ó u
U O Q 2
L
An
° ú
U
o ac
v ou L
"En
E o
n ó >
m ce
a a LN -
Ln Ó r_
Ñ
$ a
o Ñ 3
an 3 `o
Q iJ
O 2 l Z.
U N Á
u
a c
jE
TO
-r vNi eeT O Of
U c
=o
.
WL N N Ñ
>
O
Ó
~Ú
U
a O E
-
.U
ó
N
N
N
ce - -43
Q Z
Mode Selection
A selector switch on the splice logic module permits placing the machine
in the add -on, insert, or normal mode. As a special feature of the equip-
ment, the selector switch may be left in either the Add -On or Insert posi-
tion, and the machine will still make a normal recording if the RECORD
button is pressed before the PLAY button. This feature is obtained by use
of a splice -mode circuit on the splice control module; the circuit is similar
to the mode memory circuits of the control system.
When the switch is on Add -On or Insert and the PLAY button is pressed,
the splice -mode circuit becomes energized, and the machine enters the
splice play mode. In this mode, the splice circuits are placed in a state of
readiness for insert or add -on operation, as selected, and the machine will
not go from play to record unless the servo selector switch on the tape sync
processor module is on Switchlock and the machine is actually in Switch-
lock. In addition, the following operation and setup controls are actuated:
1. Aud Record Off position of toggle switch on record control panel.
2. CAP ADJ /FIXED TW /VAR TW switch and CAPST ADD ON ADJUST
control on splice logic module.
3. SPLICE RECORD HOR PHASE control on splice timing module. (If
machine is not in Pixlock and switch on splice logic module is on
Var TW.)
4. ERASE TEST buttons on splice timing module.
If the switch is on Add -On or Insert and the audio or cue record buttons
are pressed, the video circuits will enter the play condition. Therefore, the
splice -mode circuit will be energized, and, except for the audio or cue
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 299
record circuits, the machine will enter the splice play mode. When the
switch is in the Norm position, the machine cannot enter a splice mode
and therefore functions normally. Similarly, if any of the splice modules
are removed, a splice interlock circuit is broken, and the machine reverts
to normal operation.
Electronic splicing is facilitated by installation of a new ADD ON /INSERT
indicator and modification of the original SL ( Switchlock) indicator. If the
splice selector switch is placed on Add On or Insert but the splice -mode
circuit is off, the corresponding section of the ADD ON /INSERT indicator
glows red. However, when the splice -mode circuit goes on, the lamp color
changes to white. Similarly, when the servo selector switch on the tape
sync processor is on SL, the PLAY button is pressed, but the machine is not
yet in Switchlock, the SL indicator glows red; when the machine actually
enters Switchlock, the indicator color changes to white.
The momentary release positions of the toggle switches serve the same
functions as the original momentary push buttons. If the machine is in
audio or cue record, or both, and either switch is moved to the release
position, the corresponding record function will be turned off.
Horizontal- Phasing Provisions
To prevent a horizontal lurch in the picture, the phasing of the tonewheel
with respect to the headwheel during a splice recording must match that of
the original recording machine. Therefore, to permit adjusting the steady -
state headwheel phasing, a ten -turn potentiometer marked SPLICE RECORD
HOR PHASE is provided on the front panel of the splice logic module. The
control operates in conjunction with the last two positions of the CAP ADJ/
FIXED TW /VAR Tw toggle switch on the splice timing module. When
this switch is in the Fixed TW position, the phasing potentiometer is
disabled. If the toggle switch is on Var TW, the phasing potentiometer
will be actuated when the machine enters the splice play mode. The control
may then be adjusted as directed in Chapter 11 to make the new recorded
sync occur in the same places on the tape tracks as the original recorded
sync.
This test should be made only while playing back a tape containing
unneeded material, since the tape will be erased. Once the adjustment is
made, it will require no further attention until the headwheel is changed.
A short functional description of the electronic splicing system is given
in the following paragraphs. Fig. 10 -11 is a functional block diagram of the
system, and Table 10 -1 lists the major circuit designations and functions.
5 - ñó a 5:g
t?. -g cá n
l
17, E
03
CI. 00
$ Ñ
--
Ñ _á
"4'
ñ "g"
=f
O.
j
tré
s
H ó $ 1 ° â
.°S
Ñ i$
152,
0ô « o 0
Um Wh`o°
U
5 o
8
N
Splice Timing Module Bank Electronic Automatically control starting and stopping
Module Counter, Erase times of record and erase functions during
and Record formation of ingoing or outgoing splices.
Delay Circuits
Major Circuits
Mounting
Component Location Designations Functions
Indicators Mode Indicator SL Indicator Original SLindicator modified to include
Bank Above both red and white lamps. Red lamp lights
Tape Transport when machine is in splice play mode and
servo selector switch is on Switchlock.
White lamp lights when machine actually
enters Switchlock.
ADD Two sections each containing both red and
ON /INSERT white lamps. Red ADD ON lamp lights
when splice mode switch is on Add On.
White ADD ON lamp then lights when ma-
chine actually enters splice mode (splice
bus energized). Insert section operates in
similar manner when switch is on Insert.
SPOT ERASE Lights when audio spot erase circuit is
energized.
present. Actually, the master erase head is offset by about IV8 inch, so
slightly less than 15 frames are present between the heads. The offset is
provided because of mounting tolerances and circuit requirements.
Because of the preceding considerations, when the machine is set for 60-
field standards and normal tape speed, the counter is automatically set to
permit counting a maximum of 16 frame pulses just after the MASTER
RECORD button is pressed. Then, when the first frame pulse occurs, the track
on which video recording will start is exactly 15 frames in advance of the
headwheel.
At this time, the erase delay circuit is actuated. This circuit allows
enough unerased tape to go by after the end of the fifteenth frame has
passed the master erase head to create the desired overlap between the old
and new recordings ( normally adjusted for 11/2 tracks) It then causes the
.
end of the track. The splice -control switch shuts off the counter and switches
the control -track record /play module to the record condition.
At the end of the record delay ( approximately 560 microseconds ) , the
splice video record switch on the fm modulator module allows the record
current to enter the video heads. At the same time, the audio and cue splice
switches go off and allow erasure and recording on the corresponding tracks
( unless the audio or cue toggle switches on the record control panel are in
the off position) .
Recording then proceeds normally except that the guide servo is cut
out, the capstan oscillator remains in the play condition, and the oscillator
is referenced to the output of the capstan memory circuit. ( Also, if the
CAP ADJ /FIXED TW /VAR TW switch is on Var TW, the headwheel phase
is shifted by the amount determined by the SPLICE RECORD HOR PHASE
control.) This condition of the machine is called the add -on splice record
mode.
At the end of the add -on recording, the operator may immediately switch
the machine to a desired mode by pressing the PLAY, WIND, FAST FORWARD,
or STOP button. If play is selected, the splice mode circuit remains on, and
the machine returns to the splice play mode; otherwise, the splice mode
circuit goes off, and normal operation is restored.
Insert Operation
The operating procedure for insert operation is the same as for add on,
except that the mode selector switch is placed on Insert. Circuit action dur-
ing the initial phases is also the same, except for the following:
1. When the PLAY button is pressed, the control -track erase relay be-
comes energized and disconnects the control -track section of the selec-
tive erase head. As a result, the control track will not be erased during
the recording.
2. At the end of the ingoing splice cycle, the recorder enters a condition
called the insert splice record mode. In this mode, the entire capstan
servo remains in the play condition, and the tape speed is governed
by the original control track. Also, unlatching of the master -record
mode module is prevented by an inhibit voltage, and the unlatch bus
is disconnected from the play, wind, and fast -forward mode modules.
of reconnecting the wind relays so that they cannot be energized until the
capstan -roller solenoid is de- energized. Another consists of shorting the
sections of the STOP buttons in the stop ground interlock system to prevent
the mode modules from being turned off as soon as the STOP buttons
are pressed.)
No mode module for the stop mode is included in the original recorder,
but a stop memory circuit on the splice logic module serves the same pur-
pose. This circuit "remembers" that the STOP button has been pressed but
does not allow the machine to enter the stop mode until the end of the
outgoing splice cycle.
The actions occurring during an outgoing splice cycle, in 60 -field
standards and at normal tape speed, may be summarized as follows:
1. On the first frame -pulse count, the erase delay circuit is actuated. At
this time, the track at which video recording will stop is 15 frames
from the headwheel, and the point at which video erasure will stop
is 21/2 tracks closer to the headwheel (assuming that the ERASE
DELAY control was adjusted to produce an overlap of 11/2 tracks for
an ingoing splice.)
2. At the end of the erase delay period, the correct point reaches the
master erase head, and the video erase switch shuts off the erase
current. (The outgoing erase delay is automatically reduced by 4
milliseconds from the ingoing erase delay to produce an asymmetrical
overlap.)
3. At the sixteenth frame pulse (end of the fifteenth frame) , the
counter actuates the record delay circuit. After a delay of 560 micro-
seconds, this circuit actuates the video record switch and the cue and
audio erase and record switches, thereby shutting off the currents to
the corresponding heads.
4. The counter then starts another complete cycle of sixteen counts.
During this cycle, the master -record mode circuit remains on, but the
machine neither plays nor records.
5. At the fifteenth count of the second cycle, the advance transfer cir-
cuit de- energizes the video -head changeover relay and places the
guide servo in operation.
6. At the sixteenth count, the splice -control final switch stops the
counter, unlatches the master -record flip -flop, and allows the machine
to enter the selected mode. If play was selected, the machine returns
to the splice play mode. However, if wind, fast forward, or stop was
selected, the capstan -roller bus unlatches the splice mode circuit. The
normalizer circuit then restores the machine to normal operation.
NOTE: It should have been observed by the careful reader that the Ampex
and RCA systems have a different number of frames between the selective
erase head and the video head. This has nothing to do with interchange-
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 309
TR -70/60 TR -70/60
VTR VTR
to do is to enter the desired time into the unit programmer stop register
(by using the 10 -key keyboard) and press the SEARCH button. Generally,
a log of noteworthy activity is kept during the progress of these games,
and the use of the time code permits desired action to be cued reliably
without prior rehearsal.
Using the Time -Code Editor
On the Dean Martin Show, it is customary to record two separate feeds
from the originating studio. One feed carries the entire studio output, while
the second represents the unswitched output of the single camera normally
used for close -ups. Here again, the common time code supplied to both
sets of recorders permits the two pictures to be "intercut" during editing
while precise sound synchronization is maintained.
The principal elements of the Martin show are recorded on Saturday
evenings, and editing is scheduled for Monday and Tuesday. A total of
about 24 hours is required to complete the editing of the average show.
Usually, the editor and an assistant are employed in the editing process.
The editor operates the vtr on which the show is being assembled, as well
as the switching and effects equipment. The assistant handles the playback
vtr, its programmer, and the slow-motion disc equipment. When dissolves
between reels are used, a second assistant and a third vtr are assigned.
A 64-minute reel containing previously recorded control- track, black-burst,
and time -code ( from 0 to 60 minutes) signals is used as the assembly
medium. This becomes our electronically edited master, and all edits are
in real show time.
When "freeze" is employed, picture and sound from the segment to be
frozen are assembled in the normal manner on the master reel. While this
is being done, the picture is simultaneously recorded on the disc. The pic-
ture on the master reel is then replaced by a replay from the disc, with
action stopped at the appropriate point. It is, of course, essential that pre-
cise synchronism be maintained between the vtr and disc, since the disc
picture must exactly match the previously recorded sound up to the "freeze"
point. To achieve this necessary condition, use is made of a contact closure
that is externally available from the unit programmer. This closure is used
to start the disc from the master -reel time code when the material is
recorded, and, consequently, an exact time relationship between vtr and
disc is insured for playback.
Time -code -actuated contact closures are used to control a variety of ex-
ternal equipment in addition to the slow-motion disc. Sound playback
machines, auxiliary vtr's, and automatic effects units are frequently started
and switched by this means.
A and B Rolls
On the Martin show, opening and closing titles are produced during the
editing operation. Separate A and B rolls are prepared by careful assembly
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 313
from selected portions of the original recordings, and these are played back
in synchronism utilizing the synchronizing feature of the unit programmer.
The standard closing for the Martin show is 60 seconds long and consists
of roll titles matted over a series of stills followed by a segment in normal
motion. Finally, the scene fades to black. Sound during the closing is the
show theme. In a first step, as indicated previously, the editors prepare
A and B rolls of the entire series. These are rolls of tape that contain suc-
cessive show segments so arranged that, in playback, dissolves may be made
from roll A to roll B, and vice versa, and thus produce the sequence
desired. The stills required are obtained by freezing selected frames from
the original recordings. In making the assembly, a third vtr, also started
by contact closure in the unit programmer, is used to play back previously
recorded roll titles. All picture signals are routed through the switching
equipment in the room so that the roll titles may be matted and the dis-
solves executed. Theme music is also "laid down" at this time. The time -
code editing system has substantially reduced the time and effort required
to produce a sequence of this type.
if - v É
_
1(
minutes, seconds, and frames. All three units are small (approximately 5"
X 8" X 11 ") , hence the term "Mini-Modules." Basic applications are out-
lined by Fig. 10 -15. The units are completely solid state, using integrated
circuits and transistors.
Display
Monitor
14233821
Display
Some of the advantages and options available when video or audio tapes
carry the edit -code signals are:
1. Tapes can be previewed and a sqeuence log written to show start and
stop times of selected scenes.
2. Subsequent program editing, either manual or computer- automated,
can be made more quickly and at less cost than with other methods.
3. Scenes to be shown in a live program can be located more rapidly
either manually or automatically.
4. The edit code can be used to synchronize dual-equipment operations.
5. Recordings of significant events carry a permanent time record.
Sequences and single frames can be examined in terms of actual
time and with split - second accuracy.
6. The edit code can accompany the transfer of video scenes from one
type of recorder to another. For example, quadruplex recordings
can be copied by a helical recorder. The low -cost equipment can then
be used for careful scene reviewing and development of a program
sequence log. Subsequent editing of the original quadruplex tapes
318 TELEVISION BROADCASTING
.
were 30 frames per second, or 1800 frames per minute. However, at the
rate of 29.97 frames per second, only 1798.2 frames per minute are gen-
erated. Therefore, an error of 1.8 frames per minute (1800 -
1798.2 = 1.8)
+60
40
.a.,,I.im..,
IFAMIA I
+20
20
40
_60
0 2 3 4 6 7 8 9 10 11 12 13 14 15
20 30
40 50
Time (Minutes)
is induced. Because the basic video frame duration is 1/30 second, or ap-
proximately 33.33 milliseconds, the time error is 60 milliseconds (1.8 x
33.33 = 60) per minute. To overcome this error, the generator drops two
frame counts per minute for a time correction of 66.66 milliseconds (2 x
33.33). However, this effectively overcorrects the error by -6.66 milli-
-
seconds (60 66.66 = -6.66) . (See Fig. 10 -16.) At the end of the
second minute, the total error before correction is approximately 53 milli-
seconds. After the two -frame correction, the total error is approximately
-13 milliseconds. The two -frame correction is repeated until after the ninth
minute correction, when the total error is approximately -60 milliseconds.
Between the ninth and tenth minutes, the 60- millisecond error generated
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 319
7
- Ua
d
`a
d
°0
U
2p .J
m
C or \
T 1
tot a,
L t0
CD
IÓ O
a3 u
E
N O
o
j
m
CL
-O
U ..°..
L
H
°w
CC
0 u
` T CO
LL N G
N
S
O
.O
_O
r
VCO
5V
R37
BI
Cl?
Phase Comparator
841
Q9
C IS
00 X6 012
TP3
cui
R36
07
1111
R47
+I2V
111 +5V
+5V.
Phase Ref
Ref Source
Phase Comp
+12V
No Sync
No Sync Indicator
CLK 1
K 0
IC37 p
CLK
CLK
O
1/2 CB
CL
OR
ICB
ICB
CLK
CR
2C
2CB
4C
IC38 BCB
(Same as Atom/ 16CB
32C B
32CB
Reset Latch
80CB
IPPFR
61(
-d CLK
MCB
Capacitor C17 then discharges through Q10 and L1. When the capacitor
is discharged sufficiently, Q10 turns off and C17 begins charging again.
Potentiometer R37 determines the firing point of Q10 and thereby adjusts
the frequency of the oscillator.
Capacitor C17 receives an additional charge through Q9. Therefore, the
point on the ramp at which Q8 fires also determines the frequency of the
vco. Because a derivative of the vco signal is utilized by the phase com-
parator, the oscillator is self- compensating and will lock onto the refer-
ence- source signal.
32
Phase Comp
Q8 (S)
Q8 (Gli
N
I
Test Point 3
-
(Expanded)
4GOys --i
Fig. 10 -19. Phase- comparator waveforms.
Edit Code
Multitrack From Control Track BE C ?0
Audio Edit Code
Recorder Reader
Fig. 10 -20. Method of synchronizing Reference
audio recorder having control servo. Derived From Edit Code
Studio Reference
0
É
°'
c
á
-7-- " o` 'm ó v
3
i.° E
1i111ì1111
00_U
L)
03 co
V v
2O O
= C_
cc U
O
1S EN, IOS EN
OFF-ETI
Frame Rate
U o
U a
2
li
11
FRAMES and BCD SECONDS in Fig. 10 -21) , which are applied to the
frames and second display indicators and to buffers for outputs at J2.
Other signals used by the frames /seconds counter /register are (as labeled
in Fig. 10 -21) FWRD, FRAME RATE, and READ EN. A switch in the
:
frames /seconds counter /register selects the frame count (25 or 30) to
which the bcd frames are permitted to register before being reset. This
switch must be set by the operator to correspond to the number of frames
per second in the incoming edit time code.
Similarly, the minutes /hours counter /register produces parallel bcd sig-
nals (labelled BCD MINUTES and BCD HOURS) for application to the
minutes and hours display indicators and to buffers for outputs.
The HOLD switch activates a latch in the decoder which prevents any
further decoding of the ones in the decoder or changing of signals from
the counter /register. Therefore, in the hold mode, the display and output
signals represent the time and frame count at the instant the HOLD switch
was pressed. When the switch is pressed a second time, the latch is re-
leased, and the decoder, counter /registers, and display are permitted to
update to the time represented in the edit time code. The read -enable
signal darkens the frames display by disabling the high voltage to the
frames indicators.
The BE 400 Video Character Generator
The BE 400 video character generator converts any number of video
monitors into time -code displays. During video play or search, the unit
reads the edit code from the tape and generates a corresponding visual
time display on the monitor screens. Front -panel switches permit rapid
changes in the size and location of the time display, and a black back-
ground mask can be added or deleted.
Use of the unit is flexible. It can display the tape edit code (time) on
the monitors without affecting the video signals being broadcast or being
transferred to another tape. A common option, however, is to add the
visual time display to the video signal when it is transferred from a
quadruplex recorder to a helical recorder. Scene review is then easy be-
cause all scenes, moving or single frame, carry the visual time display.
The features just described make it possible to prepare, with relatively
inexpensive equipment, a program sequence log to show the start and
stop times of selected scenes. A start -stop time log is useful in a number
of operations:
r
óö
rIX
°°111.,,..
mn
, 0a
O
El
6 H
,22,35,59,11,
MOM IMIY/FS WOOS
EDIT
s..wa.
L 7C C.JI
o O®
Courtesy Electronic Engineering Co. of California
Fig. 10 -22. Keyboard of BE 210 programmer.
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 329
Controls are also provided to advance or retard the capstan servo con-
trol circuit manually.
Additional front -panel controls provide the following functions:
PROGRAMMER ON: Activates control functions of the BE 210 to con-
trol associated video recorder.
PROGRAMMER REMOTE: Delegates control of BE 210 to remote point.
PLAYBACK /EDIT: Six mechanically interlocked buttons.
PLAYBACK CLOCK: The code at the recorder cue-track input is selected.
The time code generator (clock) is normally fed to this point. This
selection allows the transport (play and stop) to be controlled from
the time -code -generator time.
PLAYBACK SEARCH: The video -recorder cue -track output is selected.
The code at this point is one of two. During standby or record, the
code at the cue -track input appears at the cue track output through
the electronics -to- electronics (E -E) circuit. When the cue track is in
playback, the code is reproduced from the tape. This position is nor-
mal for tape search and cue. In this mode, the stop time is enabled
during recue. The tape may be searched and stopped or cued to any
point. A clock or slave code may then be used to place the recorder
into play (start) When the recorder is recued, the tape will be re-
.
to be made using the edit button (after play) and an outgoing edit
to be made by operating either play, recue, preview, stop, rewind, or
fast forward. Operation of the manual button also resets the edit se-
quencer to the off condition (no arrows lighted, edit sequence but-
ton dim) .
START DISPLAY and STOP DISPLAY: Operation of the start (stop) dis-
play button transfers the time in the start (stop) storage register
into the display logic and displays this time. The time in the store is
not altered. This allows stored times to be read and modified. The
display is then restored to normal operation by one of three methods.
Operation of either the START ENTER, STOP ENTER, or NORM button
transfers the display back to the code input selector.
NOTE: The time in one storage register may be moved to another
storage register without destroying the time in the first. The NORM
button transfers the display to normal operation.
HOLD: The HOLD button freezes the display at the instant of opera-
tion. This feature is useful in displaying time without time entry.
The display is restored to normal operation with the NORM button.
The displayed time may be entered into the start or stop store if
required. Either entry button restores the display to normal.
Status Indicators- Indicators are provided on the upper left section of
the BE 210.
SYNC: Lights showing that the synchronizer has synchronized local ma-
chine to remote code.
No EDIT: Indicates edit was inhibited.
PREVIEW: Indicates that the video and /or audio monitors have switched
to the incoming program.
RECORD: Indicates that the video and /or audio monitors have switched
to the incoming program and the machine is recording.
START READY: Indicates that the start time function is enabled.
NOTE: Further editing and animation equipment involving video
disc recorders is covered in Chapter 14.
EXERCISES
Q10 -1. Why is it not possible to record a new scene into an existing program
simply by pressing the record button at the desired time while playing
a tape already bearing recorded material?
Q10 -2. Describe the basic difference between (A) the insert mode of splicing
and (B) the add -on (assemble) mode.
Q10 -3. What causes a momentary horizontal "lurch" or "bobble" in the pic-
ture at a splicing point?
Q10 -4. What controls the capstan servo in the recording machine (machine
on which editing is being accomplished) during (A) the insert mode
and (B) the add -on (assemble) mode?
Q10 -5. Basically, what is "programmed electronic editing "?
CHAPTER 11
11 -1. TERMINOLOGY
The terms most commonly used in tape production techniques may be
defined as follows:
332
QUADRUPLEX TAPE SYSTEM OPERATIONS 333
Cue -In: Method used to enable the operator to stop the tape at the
proper point prior to the start of actual program content. Visual or aural
cues may be used; the aural portion is either on the regular program
sound track or on the cue track. One common visual method is the tele-
vising of a second -clock with the second hand started about 15 seconds
prior to the program, turning counterclockwise toward zero. The signal
is brought to black level 2 seconds before program start to provide a
clean take interval for the video switcher.
In addition to all of the preceding factors, the control -track signal phas-
ing relative to the reference pulse (hence recorded video tracks) must be
standardized. A spliced -in section which is different in control -track phas-
ing relative to the preceding and succeeding sections will need to be man-
ually retracked upon entering and leaving the spliced portion. Even under
QUADRUPLEX TAPE SYSTEM OPERATIONS 335
--100
1111111 11111110
1/111111
1111111111II
I
11
II I II
111111
.;
I
Ail 1:
1111111111111110 11 I111111110,1,11
1' l
11111 II111II I I
1
II1111111179Ih Ill
!;0
II;Ij 1' Ili I Il 1 Ili Ili '!i, ,i
100 A
I I
- 100
80
Type 524 scope. This should be made evident as the discussion continues.
The recommended Tektronix scopes for modern systems and test tech-
niques are Types 545B, 547, or 453 with Mod 127C.
Thus, the "external standard" will be a scope with a 35- to 60 -MHz
bandwidth. Use this at periodic intervals to observe the terminated input
signal to the monitoring cro. The two should show the same response.
Use the external cro periodically to check the test signal input (actual
system input) across the input termination. This should be done for all
signal paths feeding the tape system and is particularly important for
color tape systems.
Use the multiburst test signal for all these checks. Then switch the
monitoring cro to the output and check the response of the E -E cir-
cuitry. On modern high -band recorders, you should see exactly what you
see at the input. On modern recorders in the low -band E -E mode, you
will normally see the last burst (4.2 MHz) slightly rolled off, with a
predominant "beat" pattern (moire) .
20 4 1 1
100 IEEE
40 IEEE
You should check the remaining response positions of the cro, as shown
by Figs. 11 -1B through 11 -1D. The low -pass position (Fig. 11 -IC) is
quite important in the following respect. Only the dc axis remains. If
there is any "selective clipping," the axis will shift at the burst frequency
where this occurs. "Selective frequency clipping" simply means that clip-
ping occurs at one or more particular frequencies. When one side of a
sine wave is clipped and then the waveform is restored to a sine wave
by filtering circuits, the dc axis is shifted. The pattern of Fig. 11 -1C, as
well as all other patterns (except Fig. 11 -1A in low band), should look
the same at the output as it does at the input. (This still concerns the
E-E mode.)
'For a more complete development of the sine pulse, see Television Broad-
casting: Systems Maintenance by Harold E. Ennes (Indianapolis: Howard W.
Sams & Co., Inc., 1972), p. 77 ff.
QUADRUPLEX TAPE SYSTEM OPERATIONS 339
-0.250 ps = 4 MHz -
DC Axis
0.125 vs = 1 Picture
Element
a.
-h.a.d
2nd Lobe
the leading -edge and trailing -edge lobes, the h.a.d., and (with the com-
bination window and pulse) the relative heights of the pulse and window.
A rapid rolloff (almost a cutoff) just above the video bandwidth con-
cerned results in practically no effect on the amplitude, but it does pro-
duce ringing. The shape of the rolloff and whether the resulting phase
shift is leading or lagging is revealed by the distribution of ringing before
and after the pulse. See Fig. 11 -3A. A fairly rapid rolloff with phase equal-
izers incorporated to correct the resulting phase distortion will cause the
ringing amplitudes to be equally distributed before and after the pulse.
The degree and nature of the high- frequency phase distortion is measured
by the departure from this optimum response characteristic. In Fig. 11 -3B,
we have an indication of increasing delay with increasing frequency. In
Fig. 11 -3C, we have an indication of decreasing delay with increasing
frequency.
The system performance K- factor is related to the following (simul-
taneous) characteristics:
Fig. 11 -4A represents the graticule of the Tektronix Type 529 scope
with the sine- window signal displayed with normal horizontal sweep. The
limits for 2- percent K- factor are shown by dash lines, and solid lines indi-
cate the limits for 4- percent K- factor. The limits for the pulse-to -bar am-
plitude ratio are indicated at point 1. The limits on "tilt" of the bar (win-
dow top) are indicated at point 2.
Fig. 11 -4B represents the same signal displayed at 0.25H /cm with the
x5 multiplier. Note that the base line is at the 7.5- percent line, indicating
standard 7.5- percent setup.
Fig. 11 -4C represents the proper display of the pulse (only) to mea-
sure h.a.d. and any ringing amplitudes. The h.a.d. point (1) is indicated
by the termination of the K- factor lines, accounting for the 7.5- percent
pedestal used. When looking at the system output on the cro, remember
to adjust the pedestal output so that the base of the pulse is at 7.5 IEEE
units. Although the graticule is marked for a T- pulse, the 2T -pulse is
measured on the same graticule by proper adjustment of the time base.
With modern high -band systems, you should see no difference on the
Type 529 monitor at the input and output for a 2T -pulse in the E -E cir-
cuitry. The T -pulse will show a difference between the output and the
input. If proper phase correction is employed, an indication something
like that of Fig. 11 -3A will be obtained. The T-pulse will be somewhat
lower in amplitude than the window.
High- frequency rolloff results in loss of amplitude of the pulse. This
results in a widening of the h.a.d., since the area of the pulse represents
342 TELEVISION BROADCASTING
-40
2% and 4% K Factor 0.125 µsec T Pulse
100
80
60
40
2% and 4% K Factor 0.125psec T Pulse
Fig. 11 -4.
QUADRUPLEX TAPE SYSTEM OPERATIONS 343
100
ao
60
40
20
1 1 1
20 I I + I
{ 1
}
f I
I
-40
2% and 4% K Factor 0.125 µsec T Pulse
Ringing_
Period
Waveform displays.
344 TELEVISION BROADCASTING
The ringing period (Fig. 11 -4D) has the following relationship to the
cutoff frequency:
R= f 1
where,
RP is the ringing period in microseconds,
fc is the cutoff frequency in megahertz.
For example, if the cutoff frequency is 4 MHz, the ringing period is:
RP = = 0.250 to
4
Then to find the cutoff frequency (in megahertz) for a given measured
ringing period (in microseconds):
f= R 1
1. Scope
1.1. This recommended practice specifies the signals to be recorded on
a magnetic video tape for use in evaluating and adjusting the perform-
ance of monochrome video-tape recording and playback equipment on a
routine operational basis. The characteristics which can be checked pri-
marily are related to the video performance, although a cursory check of
the audio channel is included for operating convenience.
Approx 8 Active Lines With Tracking Control Adjusted to Play Head Number SMPTE
Fallowing Vertical Blanking Vert Sync on Same Head as Recorded
/////////////////////////////
Band
RCA Ampex Number
///////////////////////////////////////// 2 2 1
3 3 2
4 1 3
1 4 4
2 2 5
3 3 6
4 1 7
Raster 1 4 8
2 2 9
3 3 10
4 1 11
1 4 12
2 2 13
3 3 14
4 1 15
Fig. 11 -5. Raster makeup related to heads and SMPTE band number.
1.2. Specifically, the recorded signals on the tape provide a means for
checking the following characteristics or adjustments:
2.5. The leading and trailing edges of the window signal shall corre-
spond in shape and rise time to the sine -squared pulse specified in para-
graph 2.1.6.
2.6. Overshoot of the stairstep signal shall not exceed 5 percent of the
amplitude of transition. An exception is the trailing edge of the stair -
step (leading edge of horizontal blanking), which is limited to 2 percent
in accordance with EIA Standard RS -170 or the latest revision thereof.
2.7. Multiburst frequencies shall conform with specified values within
1 percent. Total harmonic distortion content of the multiburst frequencies
3. Recording Conditions
3.1. The video- alignment tape shall conform with applicable American
standards and SMPTE recommended practices.
Use of the Test Tape
Special care and precautions should be taken in the handling of an
alignment tape. The following quotes are from the RCA instructions for
the MI -40793 alignment tape (SMPTE) :
"Check the setting of the reel brakes. Rapid stops with improperly set
brakes can damage or distort the tape.
"Be careful to wind and rewind the tape properly; the edge of the tape
must not climb up the side of the reel.
"After long usage, short-duration, small pressure changes on the order
of 0.0001 inch (0.1 mil) may begin to appear in the tape during play-
back. However, the tape may still be used by choosing the average vacuum -
guide position."
A photograph of the monitor display of the input signal for the SMPTE
tape at the time a recording is being made is shown in Fig. 11 -7. The
bands and sections established by Recommended Practice RP10 are identi-
fied on the photograph for convenience. Fig. 11 -8 is the oscilloscope dis-
to
100- Stairsteps
9
8
80- (
7 I
40-
20-
40
20
40-
20-
0-
-Sections 6, 7, 8--
-20-
60-
40-
20-
O-
1-Section 0 Section 10
-20-
response in the playback mode. (This tape at present uses the 0.125 -µs
T-pulse.)
Then it is a simple matter to compare the record mode with the play-
back mode. Make a recording of the T -pulse and window signal from a
properly set up test signal generator. On the latest high -band color sys-
0 I 1 2 3 4 5 6 I 7 8 9 I 10
Sections
Courtesy RCA
Fig. 11 -7. Monitor display of SMPTE input signal at time of recording.
tems, the overall performance should be better than the present SMPTE
RP 10 alignment tape, which is a low -band monochrome test.
The playback waveform should be within the 4- percent K factor for
the T -pulse (on modern systems) . The actual specifications are normally
given in terms of the 2T- pulse, where the playback should be well within
Courtesy RCA
QUADRUPLEX TAPE SYSTEM OPERATIONS 351
The use of the alignment tape will allow the adjustment of the vacuum
guide to the position which will assure interchangeability between tapes
made on correctly adjusted equipment.
The contents of this tape, and hence its proper usage, may be best under-
stood by referring to Fig. i1 -10. The contents of the tape are shown sche-
matically, and brief notes regarding the intended use of the tape are given.
This drawing should be referred to from time to time by the reader as he
proceeds through the following paragraphs. (Certain of the waveform
photos are included in Chapter 12.)
The tape is divided into two major sections. In the first section, the
various alignment signals are displayed in fairly short -duration groups.
This section of the tape is intended for quick cursory examination of
recorder performance, and can be used by an operator in performing a
routine check at the beginning of the day's operation or just prior to
making a recording. The second part of the tape presents the same signals,
but in segments of much greater duration. This section of the tape is in-
tended for meticulous adjustment or correction of faults in the recorder
(Chapter 12) .
Start Black
0:00
0:20
Multiburst
0.50
Ramp + Subcarrier This sequence of signals is to permit video head
1:15 alignment and guide alignment, and to provide a
Sint Window 2T, 201 qualitative check of the playback system.
1:40
Color Bars
2:10 Tape to be stopped or rewound in this black segment
2:20 only, if the complete recording is not to be played back.
3.50
6.20
7.50
6. Stop or rewind the tape only in the black -signal portions between
recordings. Rapid stops, with improperly set brakes, can damage or
distort the tape. Observe the tape as it winds on the reel. The edge
of the tape must not climb up the side of the reel, or permanent
damage may result.
7. Take precautions to avoid erasing the tape accidentally, i.e., by
pressing a setup or record button.
NOTE: After long usage, small apparent guide- pressure changes on the
order of 0.0001 inch may begin to appear in the tape during playback.
However, the tape may still be used by choosing the average guide posi-
tion.
354 TELEVISION BROADCASTING
Use of Color rs
The pro er roportionment for a 75-percent color -bar signal is shown
in Ig. -11A. Note that this incorporates a 7-percent setup. The gray
100 - 100 100
89
80 - 77 77
72
69
.15
60 - 56
48
Tolerance . 2.5 IEEE Units 46
en Dk -pk Chrominance Levels
d 40 - 38
36
28
20 - 15
12
7
0- 1
-5
Tolerance « 10%
40
l
(A) 75% bars, not split- field.
s 53'
120
Ref White.
80
60
40
22
20
Rel Black
0
Sometimes, you may encounter a tape with a leader that has color bars
at 100 -percent amplitude. Thus the yellow and cyan peaks are at a level
33 percent over the white reference. This can be confusing on the straight
color -bar pattern of Fig. 11 -11A and cause you to set the wrong playback
level unless you can recognize proper setup. You should always use 75 -per-
cent color bars, not 100 -percent bars. The advantage of the split -field color -
bar pattern (Figs. 11 -11B and 11 -12) is that a white "flag" always exists as
a 100 -percent reference level, to indicate clearly any difference in lumi-
nance and chrominance gain.
First, become familiar with the cro display of your own station color -
bar signal. Looking at the input with horizontal-rate sweep, you should
see the waveform of either Fig. 11 -11A or 11 -11B for a properly equalized
input.
Now observe the same signal with vertical -rate cro sweep. The top tips
of the chroma (yellow and cyan) will occur at 100 IEEE units (Fig. 11-
13). Note from the drawings of Fig. 11 -11 that the bottom peaks of
QUADRUPLEX TAPE SYSTEM OPERATIONS 357
chroma (red and blue) should occur at -16 IEEE units and the burst
bottoms should be at -20 IEEE units.
Since the sweep is at the vertical rate, the chroma signal from -16 to
+100 IEEE units results in a rather heavy trace. The space between -16
and -20 IEEE units, since the burst duration is shorter than the other
signal intervals, results in a fainter trace. (The graticule illumination was
turned off for the photograph of Fig. 11 -13 so that the faint part of the
trace would be visible.)
A color tape should contain at least two minutes of color -bar recording
on the leader. This enables the operator to set the individual channel
equalizers properly for playback of the program content.
-16 Units
I- -20 Units
Fig. 11 -13. Split -field color -bar pattern at vertical rate and
with wideband response.
5. Now change the cro to the system output and again check the color -
bar pattern at the vertical rate. In a tape system, the luminance -to-
chrominance ratio will be correct with proper equalization, but the
burst may not be. This involves a separate control which must be
set for proper output amplitude.
NOTE: You can also match all four head -equalizer adjustments by observ-
ing the output of the system on a vectorscope (where automatic chroma
equalization is not used or is removed for this adjustment). Fig. 11 -15
shows the appearance of multiple dots along each vector, showing un-
equal chroma amplitudes. Fig. 11 -17 shows the proper indication for
equal chorma amplitudes from all four heads (dots superimposed so that
a single dot appears for each vector.) Since each "dot" (vector) in Fig.
11 -15 is only 1/4 normal brightness, the trace is very dim. For this photo-
graph, the graticule line illumination was turned off, making the IEEE
scale more visible than the vector scale. When all dots are superimposed,
the trace brightens considerably, as in Fig. 11 -17.
Fig. 11 -15. Vectorscope display of color -bar playback with channel equalizers
not adjusted for proper match.
back with an early machine (RCA TRT -1A) when the playback head
equalizers were adjusted for an optimum picture. In these early systems,
it was seldom possible to match response at all discrete sine -wave fre-
quencies, especially on a tape -interchange basis. The best technique at
that time was to adjust all channel compensation controls for the best
match (no banding) on the first two burst frequencies ( normally 0.5
MHz and 1.5 MHz). Most of the average picture content of normal pro-
grams (in monochrome) is in the spectrum up to 1.5 to 2.0 MHz maxi-
mum. Thus, although a test signal such as the higher frequencies in a
multiburst pattern would show some banding, the usual monochrome pic-
ture content was satisfactory.
For a color signal, there is a very different situation. The chroma con-
tent lies in sidebands of the 3.58 -MHz subcarrier. Therefore, optimum
response up through the highest frequency of the multiburst pattern (nor-
mally 4.2 MHz) is extremely important. Fig. 11 -16B is a photograph of
the playback of the same multiburst pattern as that of Fig. 11 -16A, except
that it was recorded and played back on a modern high -band color record-
ing system. For a properly operating high -band system, it is necessary to
get extremely close to the monitor presentation to observe any "noise" at
all on the multiburst signal. On modern low -band systems, some slight
moire and noise will be visible on the last burst (4.2 MHz) of the
pattern.
Adjusting Monochrome and Color ATC Error Gains
Never play the monochrome and Color ATC error gains against each
other while watching a color monitor for transients and best "effect." If
you do, you may have trouble with a spliced -in portion or the next color
tape. There is really only one proper method for each:
(A) For monochrome time -error gain: Put the guide servo in the
manual position, and while observing the demodulator output on
the picture monitor, insert about a microsecond of error. Then
switch the picture monitor to the output, and adjust the mono-
chrome time -base -error gain to correct all geometric error ex-
actly. Return the guide servo to minimum error as observed at
the demodulator output.
(B) The color error gain should be adjusted while observing the
vectorscope on color -bar playback. Adjust this gain for minimum
jitter of the yellow vector. Yellow immediately follows the time
of correction, which occurs during horizontal blanking, and is
most influenced by the correction.
Observe the yellow- vector dots and the I and Q dots in Fig. 11 -17. Note
that these dots have wider excursions in Fig. 11 -17A than is true for
Fig. 11 -17B. You will note that the proper error -gain control setting is
most evident by observing only the yellow, I, and Q vector dots. If the
color -bar signal does not contain I and Q signals, observe only the dot for
the yellow vector.
Always recheck the burst -phase setting with the vectorscope after ad-
justment of the Color ATC error -gain control.
QUADRUPLEX TAPE SYSTEM OPERATIONS 361
When you do not use special effects or other mixing equipment, your
only concern in phasing is burst phase. The best way to adjust this (other
than using a vectorscope) is to have a color monitor which can be
switched between the regular program -line output and the tape- system
output. When the monitor is adjusted for proper reproduction on the
program -line output, switch it to the tape- system output and adjust the
tape burst -phase control for proper color.
Tape- Signal
Burst
System Subcarrier
Reference
Subcarrier
Amp
\ Color Error
Detector
Gated Horiz
From Sync Logic
System Phase
1Phase Error
Burst Phase
Error Signal
Color
to Colortec or
Processor
Color ATC
Burst Subcarrier
1
Signal Chroma
Information
1Sidebandsl
.r //
i
Burst Burst
Normally this is done on the color -bar leader of the tape to be played.
It is nearly always necessary to readjust slightly for proper flesh tones in
the program content. The burst-phase setting is rather broad on color
bars (unless a vectorscope is used) compared to the much more subtle
phase adjustment required to obtain good flesh -tone reproduction on a
color monitor.
QUADRUPLEX TAPE SYSTEM OPERATIONS 363
At the start of the operating day, before the tape is threaded on the
transport, ample time should be scheduled for the operator to carry out
the following procedures. (The cleaning procedure should be exercised
before every recording and before playback of programs more than 5 min-
utes long. Once a day is normally sufficient for the other procedures.)
Cleaning
Naphtha, Energine, and Freon TF are the most widely accepted clean-
ing agents for video and audio heads and all other metal contacts on the
tape transport. Chlorothene may also be used, but only with the utmost
caution. This agent (although cheaper than Freon TF in bulk quantities)
will soften the oxide binder of the tape if ample time is not allowed for
it to dry thoroughly. Denatured alcohol is also used for all items other
than audio and video heads.
Use lint -free tissue (such as Kim -Wipes) or a soft brush (camel's hair)
moistened with the solvent to clean the following items:
A. Supply tension arm
B. Supply idler
C. Video (master) erase head
D. Rotating -head panel (headwheel or drum, pole tips, vacuum guide,
nylon bearings under vacuum -guide block, vacuum -guide screw
contact with the movable servo arm, control -track head)
E. Audio and cue heads (and simulplay audio and control -track heads)
F. Take -up idler
G. Capstan and capstan pinch roller
H. Counting roller
I. Take -up tension arm
J. Take -up and supply reel hubs
The video heads should be cleaned by hand -rotating the shaft while
holding a moistened tissue against the heads and rotating with a slight
circular motion. Use the agent on the tape -supporting surface of the
vacuum guide while the vacuum pump is working, and inject a little of
the agent into the slots with an eye dropper or hypodermic injector. In-
ject a slight amount into the slip -ring assembly (on RCA heads) on the
head panel, and rotate the shaft by hand to clean the brush assembly. In-
spect each brush to be certain that the slim spring contact arm is prop-
erly seated in the brush groove. A cotton applicator or tissue dampened
with a solvent is handy for cleaning the control -track head.
Degaussing
As a precautionary measure, the tape transport should be demagnetized
at least once daily. Fig. 11 -19 shows the usual tell-tale indication in the
364 TELEVISION BROADCASTING
switched rf output at the switcher unit when magnetic paths exist. A tape
which was originally good can be harmed by this condition.
The small hand degausser which is normally purchased with the equip-
ment should be energized at least 4 feet from the tape transport and
brought slowly up to the panel. Keep the unit slowly rotating in a small
circle while going over the entire tape path from supply reel to take -up
reel; be sure to include the master video erase head. Start the video heads
rotating (push the OVERRIDE switch on the Ampex machine or the
STANDBY or SETUP switch on the RCA machine) as the degausser is
passed over the rotating-head panel. Degauss the audio heads with the
conventional audio-head unit. Slowly pull the degausser away from the
panel and de- energize after the unit is at least 4 feet away.
The fact that no magnetic tools, such as screwdrivers, wrenches, etc.,
should be stored near the transport or tape is often forgotten. Always de-
gauss tools before using them on the equipment or storing them in
nearby shelves.
When it is certain that a 1 -volt peak -to -peak level exists at the input,
the deviation control for fm modulation should be set by the means pro-
vided in the system. For example, Fig. 11 -20 shows how the deviation
control is adjusted on the RCA TR-70.
channel output, back the vacuum guide away from the heads, by
means of the electrical tip-projection control with the guide servo
placed in the manual mode of operation, so that contact exists only
Reference Pulse
(B) Underdeviation.
Reference Pulse
_ Video Higher Than
Reference Pulse
(C) Orerdeviation.
Fig. 11 -20. Demodulator output displayed at vertical rate.
QUADRUPLEX TAPE SYSTEM OPERATIONS 367
at the edges of the tape. Fig. 11 -21 shows a typical pattern obtained
under these conditions. If the guide is perfectly concentric to the
pole -tip rotation (the same pressure at top and bottom of tape) , the
height of the pulses will be equal. Adjust the height- adjustment
screw until the pulses are as nearly equal as possible. Then return
the tip -projection (guide position) control to normal.
A. Place the vacuum -guide servo in the manual mode of operation. With
the standard alignment tape threaded, place the system in the play
mode. Remove all high -frequency post- emphasis (where used) to
avoid possible geometric distortions of the vertical lines from trans-
ient effects on leading and trailing edges.
B. Set the guide -position (electrical) control on the control panel to
the position designated by the manufacturer for standard.
368 TELEVISION BROADCASTING
C. Adjust the penetration -range adjusting screw on the head panel for
straight vertical lines. (See Figs. 2 -14 and 2 -16 for time errors result-
ing from improper horizontal positioning of the guide.)
Keep the time -base concept of this velocity error in terms of the picture -
signal components. This will serve well in the future. We will discuss now
one practical application, determining pole -tip projection remaining on a
head.
Thread the SMPTE alignment tape on the machine. Observe the picture
at the demodulator output. Place the guide on manual operation (remove
automatic circuitry) , and operate the guide -position knob so as to move
the guide away from the head.
See Fig. 11 -22A. The highest- frequency burst packet is at 4.2 MHz. This
is sufficiently close to 4 MHz to make a point. If the guide is backed off so
that the first bar of band B is aligned with the fifth bar of band A (4 cycles
displacement) , the time displacement is 1 microsecond. (One cycle at 4
MHz is 1/ (4 X 106) = 0.25 µs, and since the displacement is four of these
cycles, the error is 1 µs.
With this same adjustment, observe the 2 -MHz burst. Since one cycle
at this frequency is 0.5 µs, the 1 -ps error will show alignment with the
third black line (2 cycles displacement) . There will be a slight difference
because the 4 -MHz reference is actually 4.2 MHz.
Now, if you were able to measure the actual radius of the pole -tip arc,
you would find it 1 mil too large. The rule of thumb is that 1 mil = 1 µs,
0.5 mil = 0.5 µs, etc.
Fig. 11 -22B shows the same amount of displacement, but playback is
on a headwheel with slightly less than 1 mil of tips remaining. Note that
the 4.2 -MHz sine waves are "breaking into black," indicating a loss of
signal at this frequency. A headwheel assembly is normally returned (ex-
changed) for a new assembly when the tips reach 1 mil.
Still another timing check you can demonstrate to yourself with the
SMPTE alignment tape is shown by Fig. 11 -23. Observe the demodulator
output on the monitoring oscilloscope while you adjust the guide farther
away from the head from zero alignment error, on the portion of the tape
containing the sin2 pulse.
Note that both the sync pulse and the sin2 pulse move to the right in a
faint trace. For example, assume you are using the Tektronix Type 529 cro.
The time base is set for 0.125H, and the 5 -times multiplier is used. Thus
the sweep rate is 0.025H /cm, or (0.025) (63.5) = 1.59 p.s /cm. So a 1 -µs
displacement is 1/1.59 = 0.62 cm (approx) This means that the faint
.
trace at the trailing edges of sync and the sin2 pulse will move to the right
by 0.62 cm when the guide is moved farther from the headwheel. This will
occur at the same time that the picture monitor shows the first cycle of
band B aligned with the third black line (second cycle) of band A on the
2 -MHz burst packet (Fig. 11 -22) .
QUADRUPLEX TAPE SYSTEM OPERATIONS 369
Cycles 0 1 2 Cycles 0 2 4
Displacement Displacement
Band A
/
/
ll(l
/
r7/
/i////////
IE/
////////////
1/y,,,:
/////0///////il.
7/////://///////,.....
7Ì//I //I/fl////Í
, -
Faint Traces
Many heads can be used down to about 0.8 mil before discarding, but
only at the risk of excessive clogging, noise, and lowered resolution. Also,
the recording- current drive (in the record mode) becomes rather critical
in "head optimization" for proper recording and playback ( next section) .
To avoid any confusion in your mind, answer the question, "what is
standard tip penetration ?" The tip penetration is equal to the tip projection.
This is to say that if the vacuum guide is adjusted to minimize skewing,
when playing back the SMPTE standard alignment tape, the penetration
of the video tips into the tape (tip penetration) will be equal to the video
tip projection. Therefore, the tip penetration is greater with a new head
than when this head is worn down. There is a greater rate of tip wear with
maximum tip penetration. At the same time, the greater penetration serves
as a "self- cleaning" agent for heads (minimum clogging from tape oxide
particles) and gives greater freedom from dropouts.
1. The distance from the axis of rotation of the heads to the center of
curvature of the vacuum guide is 1.0334 inches.
2. The nominal tape thickness is 1.4 mils.
3. The minimum radius of rotation of the magnetic -head pole tips is
1.0329 inches and the maximum is 1.0356 inches.
QUADRUPLEX TAPE SYSTEM OPERATIONS 371
Surface of Tape
li
3. i
3.
3.
3. 1
2. 3
2.
Tip Projection Center 2.25 Mils
2. 1
2. 0 1
1. 8 I
1. 6
1. 4 I
1. 2
1. 0 t
Quadrature Adjustment
Electronic delay correction is employed in both the record and play
modes by RCA. Ampex supplies a delay equalizer as optional equipment
for the play mode only. In the following procedure, Steps -A, B, and C
apply to Ampex; Steps D through J are for RCA.
NOTE: On the latest Ampex and RCA systems, quadrature adjustment is
strictly a factory procedure in headwheel assembly. The description here-
with applies only to older systems, a number of which are still in use at
the time of this writing.
with the alignment tape after each adjustment. The head which
corresponds to a given channel can be identified by marks scribed
on the Bakelite hub of the slip -ring assembly just forward from the
terminals where the head leads are connected.
NOTE: Where the playback -delay amplifier unit is employed in the
Ampex system, playback adjustments may be made with this unit, and
the relative calibrations of the individual channel knobs may be read for
a direct indication of which segments need to be adjusted and the direc-
tion of adjustment. This unit also permits proper playback of any tape
made from another source which may have quadrature errors relative to
the Ampex head.
Steps D through J apply to RCA machines.
D. Place the guide -position switch on the control panel in the automatic
position.
E. Install the alignment tape, and place the tape recorder in the play
mode.
F. Adjust the control -track phase control to produce the optimum
picture. Set the four high- frequency compensation controls to zero.
G. Set all four delay knobs on the playback -delay amplifier to zero, and
observe the picture. The presence of steps in the vertical lines (Fig.
2 -19) indicates that the heads are not exactly in quadrature (90°
apart) Identify the head which produces an average amount of
.
QUADRUPLEX TAPE SYSTEM OPERATIONS 375
H. The quadrature relation of the heads is the same for record or play-
back, except that the relative sense ( lag or lead) is reversed. Con-
sequently, the delay controls on the record -delay amplifier should be
adjusted as follows:
(1) Note the channel on which the delay control is set to zero on
the playback -delay amplifier. Set the delay control for the same
channel to zero on the record -delay amplifier.
(2) Turn each of the other three delay knobs on the record -delay
amplifier to the same numeral as that of the corresponding play-
back delay knob, but in the opposite direction from zero. (For
example, if the delay knob on the playback-delay amplifier is
set to +3, set the delay knob on the record -delay amplifier to
-3.)
I. Remove the alignment tape from the machine.
J. Record a test pattern or other test signal with vertical lines; play-
back of this recorded signal should produce vertical lines. In addi-
tion, when the control -track phase knob is turned to another track
position, the vertical lines should contain a minimum of horizontal
displacements. Large displacements indicate that the corrections
should be rechecked.
NOTE: The video heads can be self- checked for quadrature quite simply,
but only after head optimization, as described in Section 11 -6, is followed.
The reason for this is that one head that may be overdriven with fm
signal in the record mode will play back with a slight effective quadrature.
The procedure is to make a recording of vertical lines and play back with
the tracking control adjusted sequentially through all four head -tracking
positions, while the operator watches the playback monitor. For example,
if head 1 recorded vertical sync, and tracking is adjusted so that the same
head plays back vertical sync, no quadrature error will be apparent even
though it may exist. However, if tracking is adjusted so that each of the
four heads picks up the track recorded by the adjacent head, quadrature
error between any two heads will be revealed.
Fig. 11 -26A illustrates the playback of a test pattern with proper deviation
and head optimization in the recording process. Fig. 11 -26B shows break-
out into black of white areas following black -to -white transitions; this can
be caused by overdeviation of the modulator (excessive level) with prop-
erly optimized heads, or by normal deviation but with one or more heads
overdriven with fm signal. (This effect can also be produced by improperly
adjusted playback -channel equalization controls.)
The result shown by Fig. 11 -26B is inevitably called overdeviation, even
though this is not necessarily the case. In fact, this photograph was made
from the playback of a tape with the test pattern recorded with exactly the
same deviation as that used for the tape of Fig. 11 -26A. The difference
is in the levels fed to the heads during the record mode. Since the optimum
record currents change with head wear (heads may wear unequally) , opti-
mization procedures must be carried out at frequent intervals. Wear is
greatest during the first hours of operation of a new head assembly; there-
fore, checks should be made approximately every 10 operating hours for
the first 30 hours, then after every 20 hours of use. Poor signal -to -noise
ratio or banding (bands or unequal color and contrast) usually indicate
the procedure is past due, regardless of hours.
When heads wear, there is less shunting effect of the pole tips, and less
signal is required for optimum recording. This is the reason why tapes
sometimes play back with what appears to be overdeviation, even though
normal video levels are maintained. Tapes which are very objectionable in
this effect can sometimes be greatly improved by removing all high -
frequency equalization (where used) in the playback channels. The result-
ing picture may be soft (lacking in fine detail) , but such softness is
normally less objectionable than the severe white -level noise. On a color
signal, this procedure cannot be carried to the point where there is in-
sufficient chroma and burst level for proper handling in the color processing
system.
Video -head optimization procedures have become well standardized
as follows:
NOTE: With regard to the rf patterns of Figs. 11 -27A and 11 -27B, the
level controls should be adjusted for equal levels as in Fig. 11 -27A if the
pattern of Fig. 11 -27B is observed. For the pattern of Fig. 11 -27C, the
point of observation is before agc action, and no manual gain controls
are provided. For optimization, the individual head record currents are
still adjusted for maximum level of each individual head on playback.
Actually, the record current should be increased just slightly past the
point of reaching maximum playback level. (In Fig. 11 -27C, the rf enve-
lope is chopped to ground reference to indicate each individual channel.)
378 TELEVISION BROADCASTING
Fig. 11 -29. Waveform- monitor display Fig. 11 -30. Monitor display of poor
showing proper phasing of control- signal -to -noise ratio caused by
track signal with edit pulse. slight mistracking.
splice. (On the RCA TR-70 system, the monitoring cro presentation
of the record control track shows the edit pulse inverted, but proper
polarity actually exists at the head.)
2. For playback of any tape, the guide servo is normally placed in auto-
matic control so that skew is eliminated. If the automatic circuitry
should fail to work and time is not available to repair it, simply place
the servo in manual operation and adjust the position (tip-projection)
control to eliminate skew at the demodulator output, so that the
burden on the ATC system is minimized.
were quite "loose," particularly without time -base error video correction
circuitry.
loo
r
Bounce
o 0
110
Free -Run
(C) Free -run error.
Error
There is a way you can check your system to differentiate between play-
back stability and jitter in the initial recording. Think of it this way: If
you place the picture monitor on external sync (local sync) and play back
the tape in a nonsynchronous mode ( tonewheel or normal) , then you
observe the total jitter of recording and playback. Maintaining this same
type of operation, superimpose the tonewheel or stability dots onto the
picture monitor. These are pulses from the head tachometer; therefore the
motion of these dots relative to the picture itself is an indication of play-
back jitter only. Now put the picture monitor on internal -sync operation.
This means the sync from the video -tape playback is separated and used
to sweep the monitor. Therefore the playback jitter is taken up automati-
cally in the monitor, and the tonewheel or stability dots will now indicate
QUADRUPLEX TAPE SYSTEM OPERATIONS 383
recorded jitter only. It makes no difference whether you are operating the
tape system in a nonsynchronous or synchronous mode.
Caution: Some systems automatically switch the picture monitor to
external sync in going to the synchronous mode of operation, so it is best
to leave the tape system in the tonewheel or normal mode if you are not
sure. But check this for your particular system as soon as possible so that
you know your system.
The tachometer dots superimposed on the picture should never be so
"rapid" as to cause a "double- dots" appearance. If this shows up, either in
playback jitter or record stability, the chances are that you will not obtain
synchronous operation. A wide excursion (several inches), if at a slow
rate, is not an indication of jitter.
Obviously if this jitter shows up only as a recorded jitter, you cannot do
much to save the program, except to play it in a nonsynchronous (mono-
chrome) mode. If jitter shows up in the playback check, the servos need
attention. There is no substitute for the manufacturer's instruction book
here, but Chapter 12 will help give you a better understanding of servo
setup and troubleshooting.
After you have matched the head as closely as possible to the tape in
nonsynchronous operation (minimized errors as much as possible), you
can then try synchronous operation (with ATC or TEC on) If you can.
now achieve lock -up, readjust the individual channel equalizers and go on
from there.
The important thing to know is what was just described, how to tell
whether a particular tape is suitable for synchronous operation. You also
need to be able to tell whether the servos are at fault (playback jitter) or
the fault is in the particular tape recording involved.
Some erratic conditions, such as video tear -out on a certain head or heads,
can often be blamed on improper head antiresonance adjustments in the
latest systems where this feature is included. This adjustment (or adjust-
ments) compensates for the resonance of the head during playback and is
made by operating reactance (usually a capacitor) and resistance controls
on each channel playback amplifier while observing a sweep pattern (in-
jected by a special probe on the headwheel) at the fm switcher output. For
example, the procedure for the RCA TR -70 is as follows:
1. Place the machine in the stop mode, and remove the tape from the
headwheel panel and audio heads (rewinding the tape is not neces-
sary) .
2. On the switch panel just above the modules, set the TEST SELECT
switch to Res (resonance) and the CHANNEL SELECT switch to 1.
3. Remove the head resonance probe from the HD RES PROBE socket on
the switch panel above the modules by rotating the probe clockwise
approximately 45° and pulling the probe out. (Later RCA head -
wheels have a built -in probe.)
384 TELEVISION BROADCASTING
4. Insert the extruded portion of the head resonance probe into the
slot provided for it on the headwheel panel, so that the U- shaped
wire loop is close to the headwheel.
NOTE: The HD RES PROBE socket contains an interlock switch that pre-
vents the machine from being placed in any operating mode when the
head resonance probe is removed from the socket.
5. Push the TEST switch /indicator on the switch panel above the
modules.
6. Press the FM LEVEL button on the cro switcher panel.
Reactance and Resistance Controls Set Correctly Reactance and Resistance Controls Set Correctly
Reactance and Resistance Controls Set at Minimum Reactance and Resistance Controls Set at Maximum
NOTE: The smooth or linear portion of the waveform may slope in either
direction. Proper adjustment results in the desired smooth portion regard-
less of direction of slope.
1. With the machine in the master record mode, the control -track play-
back button on the monitoring cro is depressed to observe the
386 TELEVISION BROADCASTING
i
Shoulder
rv\
Flat Shoulder
15
Correct
Frame-Pulse 10
Amplitude
Frame Pulse
To setframe -pulse
phase correctly,
reduce amplitude
to very low value.
O
Fig. 11 -34. Control -track record waveforms for RCA TR -70 recorder.
NoVertical Line
Displacement
8. Go into regular splice operation. The REC button must be pressed 1/2
second before the splice is to be made. Then the PLAY button must
be pressed 1/2 second before the end of the splice. Remember that
you must go into the play mode before the record mode; otherwise,
the machine will make a regular recording, ignoring the splice mode.
NOTE: This section has discussed only the basic technique of electronic
splicing. The use of more refined automatic editing programmers varies
considerably with every make and is beyond the scope of this book. The
operator familiar with Chapter 10 and this chapter will be able to master
those techniques more quickly by following the operational outlines ac-
companying the equipment.
EXERCISES
Q11 -13. In either the RCA or Ampex high -band tape systems, the video
processing amplifier can be operated in one of two modes, the
external- reference mode or the regenerated -sync mode. What is the
difference?
Q11 -14. Does a "standard test tape" serve as a standard for both new and
worn heads?
Q11 -15. What is the minimum tip projection normally allowed on heads
before they are returned to the factory for exchange?
Q11 -16. Is it advisable to reduce tip penetration below the "standard" to ob-
tain greater life of the headwheel?
Q11 -17. How do you lap- dissolve between a studio camera and a tape
machine or between two tape machines?
Q11 -18. What can be wrong if no movement of the tape occurs in a play
mode but everything is actually working properly on the machine?
CHAPTER 12
CAUTION: New tape from the factory will often exhibit what seems
to be an excessive number of white flashes (dropouts) when the first re-
cording is played back. When tape is first received, a spot check should be
made (recording sync only for about the first 5 minutes, 5 minutes near
the middle, and 5 minutes near the end of the reel) to observe the playback
raster. If dropouts are numerous, the tape should be polished by allowing
the entire reel to run through the heads two or three times. A worn head
is best for this polishing operation. This is not quite so important when a
dropout compensator (Chapter 8) is employed.
5. The playback signal -to -noise ratio of a recording made by the system
concerned. This is an excellent indicator of any system degradation
other than amplitude- frequency response. ( Amplitude- frequency re-
sponse is usually revealed by evaluation of the composite video signal.
The maintenance engineer should bear in mind that any standard test
tape such as the one described in Chapter 11 is a check of the playback
function only. For overall record -playback checks, it is necessary to have
access to the necessary test -signal generators. Conventional signal -analysis
techniques are used when dealing with individual test signals. The test -
signal input to the recorder must be a standard composite signal. Some
older test -signal generators contain only horizontal sync so that clamping
circuits will function properly. Fig. 12 -1A is the vertical -rate display of
the output of a typical older stairstep generator; this display indicates that
horizontal pulses are contained in the vertical -sync interval. A television
tape recorder will not make a satisfactory recording of this signal. The same
only integrated, but they are also effectively differentiated so that a par-
ticular time reference is available for generation of the reference pulse.
When a signal such as that in Fig. 12 -1A is fed to the system in the record
mode, no reference pulse is generated (or if it is generated, it will be
erratically keyed) , and playback may be either erratic or impossible.
NOTE: See Section 12 -4 for special test procedures employing noncom -
posite signals.
39 -ï TELEVISION BROADCASTING
it is possible to check the heads for these conditions; such checks should
normally be made by the maintenance engineer immediately on receipt
of a rebuilt head assembly.
Check for Azimuth
A head with gap tilt lays down a narrow track, as shown in Fig. 12 -2A.
This results in a loss in amplitude, bandwidth, and signal -to -noise ratio.
An exaggerated azimuth error is shown in Fig. 12 -2B to illustrate how the
recorded track is narrowed. This is equivalent to widening the playback
gap of a normal head. The tilt of tracks recorded when the gap is not
parallel to tape travel is shown in Fig. 12 -2C, and the method used to
detect this error is given in Fig. 12 -2D. Make a recording of vertical lines.
Playback should be done with the tracking adjusted so that the same head
that plays vertical sync also recorded vertical sync. When tracking is per-
fect, the head gaps are directly over the recorded tracks. By slightly mis-
tracking on each side, any apparent quadrature effect will reveal azimuth
error. Mistracking in the direction of 1 (Fig. 12 -2D) will retard the band
in time, and mistracking in the direction of 2 will advance the band in time.
If the lines are parallel to tape travel, no error will occur. The head assem-
bly must first be optimized and aligned properly, as outlined in Chapter 11.
Parallel to
Tape Travel
Rotation
1 Gap
- - 10 Mils
Tape
I I
II ll
I l
Time Difference
_1
Head Gap
¡Rotation
1deoTracK
Head
Gal
rTape 01 i
Head Mistracked in Other Direction
I
ÌJ
Advanced in Time
banding results when the pole tip scans properly across a portion of the
track, but the remaining portion of the track is curved relative to the
standard track. Thus, a uniformly high noise level throughout one band
indicates that one head is not in the same vertical plane (axial position) as
the other three heads. Half- banding results when the plane of pole -tip
rotation is not perpendicular to the tape. In either case, the manufacturer
will allow full rebate on such a head assembly if it is returned within the
first 10 hours of operation. (Check your manufacturer on this, as this "free
time" may be lengthened.)
Tip Ringing
In color, if three or four bands of lines (indicating one of the four video
heads) show excessive moiré, this may be caused by tip ringing. Assuming
the playback level and equalization for the particular head in question
have been properly adjusted, tip ringing can cause a color interference
pattern (moiré) even when not very noticeable in monochrome.
Tip ringing is a head pole -tip fault caused by deterioration of the
magnetic -gap material. The tip can actually dig into the tape in a severe
instance of this condition and cause excessive tape shredding with possi-
ble tape damage.
Thus, cross talk from the 900 -Hz output of one of the counters results in
14 displacements, which is quite close to the 16 -band ( 960 Hz) raster con-
struction of a tape-recorder output. Sometimes a scallop effect, which can
cause the operator to misadjust the vacuum -guide height adjustment in an
attempt to eliminate an error, is actually caused by the local sync generator.
However, this effect will be observed at the input to the tape system when
cross talk exists. Certain types of master monitors employ a synchronization
which, for all practical purposes, results in line -to -line sync. This tends to
start each active line of picture information at the same spot following
horizontal sync and minimizes any effect of line displacement that normally
would be quite apparent on an average -sync monitor (and on most modern
398 TELEVISION BROADCASTING
home receivers). For this reason, the station should monitor tv -tape systems
with an averaging -sync monitor.
When "synchronous- type" cross talk is found, the indicated counter
should be additionally shielded or the wiring rerouted until the interference
is eliminated.
A more prevalent type of sync cross talk occurs between two non -
synchronous sources, such as local and network signals or local and video-
tape signals. This trouble is evident as a vertical bar or line moving non -
synchronously back and forth horizontally on the video monitors when the
network or television -tape signal is fed to the program line. (Of course,
this will not occur if the tape is operated in the Intersync or Pixlock mode,
because the tape output then is in phase with the local sync generator.)
The condition is caused by cross talk between the local sync and the non -
synchronous tape or network signal.
This trouble is usually the result of ground loops. A ground loop in an
otherwise well designed installation is most often the result of an open or
intermittent ground at one end of a coaxial cable. When an open occurs,
the signal at the open-shield end must obtain its ground return through a
number of racks. Each cable should be disconnected from the sending end
and checked with an ohmmeter from center conductor to shield to deter-
mine if the termination resistance is obtained. If the shield is open, no
continuity will exist. Always twist both the sending and receiving con-
nectors while making this check so that loose, intermittent, or high- resist-
ance connections will be made evident.
More recent sync generators, particularly those of modern digital design,
are normally quite stable and trouble -free compared to the older types. The
stability of the color -subcarrier generator is many times better than that of
older systems, and this is particularly important for second- and third -
generation dubbing on color tape systems.
Minimum input: 0.5 volt (p-p) for same signal -to -noise ratio
0.25 volt (p -p) just above sync breakout
The above situation does not hold true for modern low -band and high -
band systems, such as the Ampex VR -3000 and AVR -1 or RCA TR -70
systems. The normal multiburst test signal is used for these systems, as
illustrated in Fig. 12 -4. The extended system capability is brought about
Fig. 12 -4B shows the output of a modern low -band color tape system.
The reproduction of the full multiburst test signal by a modern high -band
system is practically indistinguishable from the input signal in a photograph.
If a video sweep is used for fm circuit alignment, remember that the
highest video frequency (4 MHz) is represented by a carrier frequency of
1 MHz (lowest carrier frequency) Thus, peaking the carrier at the lower
.
Coaxial Connectors
The pad is inserted between the test signal generator and the preamplifier
input by means of short coaxial cables. The test signal generator should be
adjusted for a signal output of no more than 0.5 volt (p -p) , to insure that
the 46 -dB attenuation will prevent any possible overloading of the pre-
amplifier.
Inability to bring a tube -type video amplifier into proper amplitude -
frequency response specifications by adjustment of the appropriate peaking
circuits may be due to any one or a combination of the following faults:
1.Changed plate load. An increase in the plate load will reduce the
response at higher frequencies. A decrease in the plate load will
reduce the gain and increase the response at higher frequencies.
2. Defective peaking coil or swamping resistor across the coil.
3. Low- transconductance tubes. (However, in certain types of negative-
feedback amplifiers, loss of highs may result from tubes that show
normal transconductance on a tube checker. Always replace tubes in
negative- feedback amplifiers before making further checks for troubles
when video-response checks indicate loss of high- frequency response.)
Always bear in mind the turn- around in fm peaking circuitry due to the
nature of the modulation system. In this case, for example, an increase in
plate load, causing reduced response at the higher carrier frequencies, re-
sults in a loss of low- frequency video and an increase in high- frequency
video.
The evaluation of low- frequency response is important in determining
absolute values of picture streaking. It is important to understand that the
degree of streaking observed on a picture monitor depends not only on the
monitor -amplifier characteristics, but also on the ratio of the brightness-
and contrast -control settings. There is almost always some visible streaking
when a window signal, or any other white bar on a black background, is
displayed if the bar extends over an appreciable portion of the scanning
line. This was the primary reason for development of the white -window
test signal, so that a truly accurate measurement can be obtained from the
cro presentation in quantitative terms.
The phase characteristic related to the low -to-high frequency response
ratio is closely related to the absolute measurement of the low- frequency
response. The shape of the passband response curve determines the tran-
sient response of the system. Depending on the rise time of the window
signal, an indication of transient response may be determined to some
degree, but this characteristic is more accurately shown by the sine pulse.
NOTE: The makeup and practical measurements of the sine- window
signal and the 20T modulated test signal were covered in Chapters 2 and
8 of Television Broadcasting: Systems Maintenance,1 and will not be re-
100 E
80
60
40
I Back Porch Level
-20
/
I
J L
I I
\ -l- 1
1
m
20 I I I 8c I I I
-lcmor0.1251
Limits -Inner: - 2% Outer: K ^ 4%
K
Window and 0.125 us T Pulse Response
...L- -40
NJNVIlN!!.
80
60
7cm or 0.175H
3cm
40
1 cm or 0.025H 3cm or 0.075H 2cm or 0.05011-
20
J L
7.5-
20
40
peated here. The student or practicing engineer not familiar with such
techniques should review this material at this time.
When the (HORIZ) MAG switch is set to X 1, one complete horizontal line
is displayed in a sweep length of 8 cm (see Fig. 12 -6A) In addition, .
refer to Table 12 -1, which lists the sweep -time relationship between H /cm
and microseconds /cm.
NOTE: The 2 Field position is the only position of the DISPLAY switch
for which a free -running trace can be obtained in the absence of a trigger
signal. Thus, the 2 Field position is useful for determining whether the
video signal is absent.
80
PO
K- Factor of Pulse
40
20
20
Limits -Inner: K - 2% Outer: K - 4%
Window and 0.125µs T Pulse Response
-40
The outer limits in each case are marked with solid lines and designate
a K- factor of 4 percent. The inner limits for each case are marked with
dash lines which indicate a K-factor of 2 percent.
Although the graticule specifies a 2- percent and 4- percent K- factor for
a 0.125 -µs T- pulse, the same graticule is used for the 2T pulse by selecting
the appropriate time base on the Type 529 monitor. A time base of
0.125H /cm X25 is used for the T pulse, whereas the 0.25H /cm x25 time
base is used for the 2T pulse. The 2T pulse (0.250 -µs half -amplitude dura-
tion) is normally used for video-tape systems measurements. Fig. 12 -7 illus-
trates the reproduction of the 2T-window test signal of the high -band test
tape mentioned in Chapter 11. For proper measurement, the sine pulse
should be adjusted by scope gain to reach the top of the bar signal (100 -
percent reference amplitude) as shown. The pulse should then fit well
within the 2- percent K- factor lines (dash lines) on a properly functioning
high -band tape system.
Top of Bar IWindowl Signal 20T Modulated Pulse WindowlBarl 2T Pulse
.o.
aa-...nM ...
d111,
dl
d2
T
12 -10A shows the crt presentation for playback of the high -band test tape
mentioned in Chapter 11. Fig. 12 -10B shows the same presentation with
the Type 529 monitor placed in the high -pass position, with additional
gain used to bring the maximum portion of the envelope to 100 IEEE
units. Thus if a = 100 and b = 95, there is 5 percent differential gain at
3.58 MHz, white compression.
IMPORTANT NOTE: A rather complete study of differential gain and
differential phase at 3.58 MHz, as well as techniques for using the vector -
scope for differential -phase measurements, was presented in Chapter 8 of
Television Broadcasting: Systems Maintenance.3 Techniques and inter-
pretations are identical for the television tape system. Tests should be run
at 10 -, 50 -, and 90-percent APL for both differential gain and differential
phase, as fully covered in the above -mentioned reference. In video tape
systems, differential gain and phase can be caused by the head or head -
resonance adjustments. In the Ampex AVR -1 system, differential gain
controls are included in the playback amplifiers.
An example of a handy form to be used for differential -gain and differ-
ential -phase measurements is presented in Fig. 12 -11. In this example,
APL
Record Playback
APL APL
On On
10% 50% 90% 10% 50% 90%
A
A B
A
B B
A
C B
Fig. 12 -11. Example of form for differential gain and phase measurements.
three record /reproducers are included in the complete system. The test
signals are first recorded on machine A, and then played back on all three
machines. The performance is recorded on the form. This process is re-
peated for all three machines. This technique gives the maintenance tech-
nician a good indication of performance of each system, as well as verifica-
tion of having all machines adjusted to a "standard" of operation.
Signal -to -Noise Ratio
One video measurement now remains: the video signal -to -noise ratio.
Measurement of recording -playback signal -to -noise (S /N) ratio is a good
indicator of the need for a complete run -through on tube replacements
(on older tube -type systems, or in modern record -playback amplifiers using
tubes), amplifier gains, input source and output lines, and demodulator
limiting circuitry. Some modern systems still employ vacuum tubes in the
record function as head drivers, and in the playback function as the first
preamplifier stages where nuvistors may be employed. All other circuitry
is normally solid- state.
Three generally satisfactory methods of measuring video-tape S/N ratio
will be described: (1) the oscilloscope method, (2) the vtvm method,
and (3) the Rohde & Schwarz UPSF video noise meter method. Method
1 is less time consuming and most convenient, but less accurate than
methods 2 and 3. Method 3 is the most accurate and is now being used
by the equipment manufacturers for factory tests.
The Oscilloscope Method-The tape used should not be completely new
and unpolished, nor old and worn or scratched; a tape should be used
that is in the condition employed for normal satisfactory recording and
playback in the daily schedule. The heads should be optimized, the de-
modulator properly balanced, and critical (optimum) tracking adjust-
ments made. To make the readings significant, the signal -to -noise ratio
should be determined on a peak -to-peak video to rms noise basis.
A. The test signal can be a "doorstep" ( three steps: one at black, one
at gray or midway between the black and white peaks, and one at
reference white level) or the standard stairstep signal described
before. Many prefer the "one- line -in -six" stairstep signal (Fig.
12 -12A) adjusted for a 50- percent duty cycle. This signal provides
a convenient method of swinging the modulation over the refer-
ence frequencies to provide the reference level, while simulta-
neously providing for a gray -step measuring point.
B. With the system optimized for recording, make about 5 minutes of
recording of the test signal.
C. With the system optimized for playback, connect an oscilloscope at
the demodulator output, and observe the playback of the preceding
test signal.
D. Fig. 12 -12B illustrates a typical playback of the test signal of Fig.
12 -12A. Note the additional thickness of the gray step. Adjust the
video output level to obtain the standard 0.714 volt (p -p) of video.
NOTE: Measurements of S/N ratio are made at any point ahead of the
ATC and processing amplifiers, to avoid false readings from regenerated
sync and spurious signals from any random ATC operation.
F. To convert the 100 millivolts from peak -to -peak to rms value,
multiply by 0.35:
100 X 0.35 = 35 millivolts
= 0.035 volt
G. The voltage ratio is now 0.714/0.035, or 20.4. The dB equivalent
for a voltage ratio of 20.4 is 26. This is on a peak -to -peak video to
rms noise basis.
H. To convert the 10 -MHz bandwidth reading to a 4 -MHz bandwidth,
which is the useful passband of the video information, add 7 dB to
the above computation:
26 dB +7 dB = 33 dB
If a 4 -MHz scope amplifier or preamplifier is used, delete Step H. The
bandwidth is given in terms of the 3 -dB -down point. Note also that from
the analysis, 100 millivolts of noise is approaching the maximum allow-
able to meet specifications for a 35 -dB signal -to -noise ratio. Thus, the
100- millivolt (p -p) noise level becomes the warning flag on older low -
band systems. Newer systems achieve an S/N ratio of 40 to 46 dB. Also,
when later Tektronix scopes, such as the Type 547 (50 -MHz bandwidth) ,
. .
. .
. .
.
MINIM /NM
.
if u te o'
I 1 .
Fig. 12 -12. One -in -six stairstep signal, 50% duty cycle.
(E) Place the machine in the reproduce mode and adjust the video
output level control on the demodulator for an output level of
0.7 volt peak -to-peak video, not including sync.
(F) Remove the video input cable from the input to the modulator
in order to disable clamping. Set the carrier frequency control
for a carrier frequency of 6.0 MHz, corresponding to a medium
shade of gray.
(G) Record a two-minute section of undeviated 6.0 -MHz carrier and
rewind the tape to the beginning of this recording.
(H) Remove the video cable from the demodulator output, and con-
nect the special bandpass filter to the output of the demodulator.
Connect the output of the filter to an rms- measuring vtvm.
(I) Place the machine in the reproduce mode; set the scope -selector
switch on the left-hand control panel to the switcher output posi-
tion, and adjust the tracking control for maximum amplitude of
the scope presentation.
(J) Read the value of rms noise on the vtvm, and compare this fig-
ure to the 0.7 -volt (p-p) level. Remember to take into account
the measured insertion loss of the filter (approximately 5 dB for
the filter shown in Fig. 12 -13) . Twenty times the log of this
ratio is defined as the video signal-to -noise ratio of the television
recorder.
NOTE: Later systems (high -band and low -band) employ built -in facilities
as described following the discussion of the noise-meter method.
47 pF 39 pF 22 pF
0.01 pF
Input
(Output of Recorder) 570
=68 pF *-1( 20pF
J To VNM
130 pF1 12P pF
75 22 p 1K
(A) Circuit.
North ill Coils Crest IMicrotranI Coils Frequency Relative Attenuation (dB)
Coil Part No. Value in pH Part No. Value in pH <100Hz 45
LI 1000 -F 14.5 -22 200 -3 6.6 -20 <10kHz 6
L2 1000 -1 55 -110 200 -5 33 -110 10kHz-4MHz 0
L3 1000 -H 29 -55 200 -4 16 -55 4.2 MHz 6
L4 1000 -F 14.5 -22 200 -3 6.6 -20 4.5 MHz 40
Cancelling Pulses
7I
I i
Compensation I i
i
i
Low-Band Color 40 dB
High -Band Color 46 dB
'Ratio of p -p noncomposite video to rms noise; noise measured at midgray level using wide
band true -rms voltmeter through 5.8-MHz low -pass filter; standard tip engagement.
(A) Set the AFC /CLAMP switch at one of the gray -level positions (i.e.,
position 3 or 4). Turn the VIDEO INPUT LEVEL switch to the off
position. Set the SERVO MODE selector to the Preset or the Line
Lock position. Thread a blank tape on the tape transport.
(B) Initiate the record mode, and record the gray -level frequency for
2 or 3 minutes; then press the STOP push button.
(C) Turn the TRANS SUPP switch (on the setup control panel) to the
CW setting. Connect the video-signal system output through a
5.8 -MHz low -pass filter to an rms- indicating meter. The rms
meter will indicate the E -E signal -to -noise ratio, which should be
60 dB or better. Reproduce the recording made in step B. The
rms meter will now indicate the rms noise content of the repro-
duced signal. The meter indication may then be compared with
the 0.7 -volt (or 0.714 -volt, depending on the recording standard)
normal level of the video portion of the composite signal, to de-
termine the S/N ratio.
NOTE: The CW setting of the TRANS SUPP switch causes the transient
suppressor to produce transient -suppression pulses that overlap the
interval during which the rf switch is transferring the pickup to succes-
sive video heads. Necessarily, these pulses occur at the horizontal rate as a
function of the phase of the head drum. Thus, this facility eliminates
head -switching transients which would otherwise greatly increase the rms
noise indication and produce an erroneous measurement of signal- system
noise.
There are inevitably many potential generators of noise in a complex
electronic system. (Amplifiers and random thermal effects are examples
of noise generators.) In the VR-2000B system, generated noise is ex-
cluded from the video -signal system because its input is not connected to
the output of the video heads until they have engaged the tape. At all
other times, the video -signal system receives incoming video from the
output of the modulator. This exclusion of generated noise eliminates the
time that would otherwise be required for the video electronics to first
recover from a noise condition when the play mode is entered. Thus there
is no normal condition in the VR-2000B system under which noise is
passing through the video signal path.
RCA TR -70 S/N Ratio Measurement -The video signal -to-noise ratio
is defined as the peak -to -peak video voltage divided by the rms noise volt-
age. (Video here does not include sync. The noise is random noise only
and is not meant to include other undesired components such as switching
transients, cross talk, moiré, tape scratches, or similar items.)
The machine is first set up to record and play normally, with particular
attention on fm deviation and output video level. The normal level of the
video portion of the signal is 0.714 volt peak -to -peak for a one-volt com-
posite output (100 IEEE units of video and 40 IEEE units of sync) This.
0.714 volt is used as the video signal level for S/N calculations. The ma-
chine output is adjusted to give exactly this level as measured on an
oscilloscope.
To read the noise level separate from the video, a test recording is made
in the noise test mode, thus giving a recording of unmodulated fm car-
rier. The machine is then played back using the noise test mode, and the
demodulated noise output of the machine is measured with a meter. The
ratio of 0.714 to the noise reading gives the numerical signal -to -noise
ratio. For example, a noise reading of 0.00714 volt would give an S/N
ratio of 100 times (40 dB) .
The test recording for the noise measurement is made with the fm car-
rier at a frequency corresponding to picture gray level. The noise measure-
ment is made at a point in the system just ahead of the ATC and process-
ing amplifier. This is done to avoid false readings due to regenerated sync
and spurious signals due to random ATC operation.
In order to obtain a representative measurement of the video S/N
ratio, the operating conditions must be carefully controlled. The tape must
be a high -performance type and in good condition. The meter and oscillo-
scope must be carefully calibrated. The meter must be a true rms- reading
meter such as the Hewlett- Packard 3400A. Capstan tracking must be
carefully adjusted, and the guide position must be carefully set to mini-
mize switching transients. Another factor to consider is the condition of
the heads. The S/N ratio improves as the heads wear and the pole -tip pro-
trusion decreases. Of course, when the heads wear out, the S/N ratio rap-
idly gets worse.
If a Hewlett- Packard 3400A meter is not available, an approximate
measurement can be made with a Hewlett- Packard 400D meter if 2 dB
is added to the noise reading (subtract 2 dB from the S/N ratio). If no
meter is available, an approximate reading can be made by using the oscil-
loscope to measure the noise. In this case, peak -to -peak noise should be
measured. It has been found experimentally that, under these conditions, a
suitable conversion factor is 12 dB. Thus, divide 0.714 by the peak -to-
peak noise as seen on the oscilloscope. Express this ratio in decibels and
add 12. The noise reading should not include extreme noise peaks.
Proceed as follows to perform signal -to-noise measurements on the
RCA TR -70 system:
QUADRUPLEX TAPE SYSTEM MAINTENANCE 417
that on many Hewlett- Packard meters 0.0071 volt is nearly in line with
-41 dB. On these meters, the S/N ratio can be read directly by subtract-
ing one. Thus -41 = -40 dB.
19. If the specified S/N ratio (same as in Table 12 -2) is not obtained,
perform the following steps to effect an improvement:
(a) Use another tape.
(b) Degauss the heads.
(c) Reoptimize the head currents.
(d) Observe the noise with an oscilloscope, and check for the
presence of hum, interference, oscillations, etc. The noise
should be random.
(e) Look at the E -E noise with the meter. To do this, use the same
procedure as for record - playback noise, but with the machine in
the stop and E -E modes, set the TEST SELECT switch on Noise,
and press the TEST switch. The E -E S/N ratio should be from
-55 to -60 dB.
(f) Use another headwheel panel.
(g) Change the nuvistors in the head preamplifier (playback) and
compare the S/N -ratio measurements.
Recording and Playback of Nonsynchronous Signals
Certain troubleshooting procedures require the use of nonsynchronous
signals which are recorded and played back as necessary. These non-
synchronous signals may consist of sweep, cw sine waves, or square -wave
signals not containing sync or blanking information.
Recording Procedure -To record nonsynchronous signals, proceed as
follows:
1. Check the level indication on the cro, and set the input level at its
source to provide an indication of 0.7 volt (100 IEEE units). This
is the correct input level for recording.
2. Set the TEST SELECT switch on the fm test panel to the Input posi-
tion.
3. Press the MASTER RECORD switch on the record panel; then press the
TEST switch on the fm test panel.
Upper Limit
Lower Limit
Upper Limit
}Optimum Equalization
Adjustments, All Channels
Lower Limit
(A) To begin the procedure, record the standard test signal for sev-
eral minutes, taking care to set the luminance deviation correctly.
(B) While reproducing the recording just made, examine the repro-
duced waveform. The output of the demodulator may be exam-
ined in the display of the waveform monitor by pressing the
DEMOD output push button on the monitor switching panel. If
all the compensators are correctly adjusted, and differential gain
has been minimized, the display will be the normal rectified en-
velope of Fig. 12 -17. (The normal noise shown is inherent in
the record /reproduce process.) If the compensators are not set
correctly, differential gain will be present, and the rectified en-
velope will slope in a direction and at an angle that is directly
related to the accumulated adjustment error. The differential gain
of the off -tape signal should be compared with that of the input
signal by alternately pressing the DEMOD and INPUT push buttons
(of the video monitor selectors).
QUADRUPLEX TAPE SYSTEM MAINTENANCE 421
Peak
White
Level
Blanking Level
Fig. 12 -17. Input test signal vs output of subcarrier filter and detector
Ampex VR-2000B system.
(A) Turn all but one of the channel equalizers fully clockwise; leave
one at its normal setting.
(B) Set the TRACK SELECTOR at its Home position, and reproduce the
differential -gain test signal recording. Note the degree of differ-
ential gain of the channel selected in step A.
(C) Repeat step B with the TRACK SELECTOR set successively at 2, 3,
and 4. Determine whether or not a change of height or flatness
occurs. If each of the heads has been properly optimized in re-
cording, the differences observed will be negligible; if one of the
recorded tracks appears to differ from the others, that recording
channel should be reoptimized by the standard method previously
described. Table 12 -3 may be useful in determining which head
made the recording of a specific track.
- 1
2 2
1 4
3
2
1
3
3 3 1 4 2
4 4 2 3 1
4. Become thoroughly familiar with the normal flow path of all con-
trol pulses so that the inputs and outputs of the various individual
chassis or circuits can be quickly traced.
5. Check power supplies and power- supply regulation on a regular
basis.
6. A closed -loop servo system has two major functions, loop gain (can-
not exceed unity) and phase gain (cannot go to 180° ) If operation
.
does not remain within these limits, oscillation occurs. The load on
424 TELEVISION BROADCASTING
1. The field -rate playback signal must be made a standard color field
rate of 59.94 Hz. Therefore, there must be a comparison between
the field rate coming from the tape system and a standard field rate
supplied by the local sync generator locked to a color standard. Any
error between these two signals must be utilized to control the head
(and capstan) servo.
2. The horizontal -rate playback signal must be made a standard color
line rate of 15,734 Hz. Therefore, there must be a comparison be-
tween the horizontal rate coming from the tape system and a stan-
dard horizontal rate from the local sync generator locked to a color
standard. Any error between these two signals must be utilized to
control the head (and, in turn, the capstan) servo.
QUADRUPLEX TAPE SYSTEM MAINTENANCE 425
3. Now that the servos are being held to a basic color standard, the
"dot interlace" of the color -subcarrier information must be main-
tained in spite of the tolerances of the mechanical head-to -tape
contact. To do this, there must be a comparison between the color
sync burst coming from the tape and the color subcarrier from the
local color standard. This comparison occurs in the Color ATC sec-
tion of the video control circuitry. This comparison enables cir-
cuitry which provides a "regenerated burst" which can then be
phased for proper flesh -tone reproduction on the color monitor. If
you keep the time -base concept, you realize this is the "tightest"
servo and signal requirement for the tape system.
To Variable Delay
To Vert -Coincidence (Dependent on Coincidence
Detector and TVA Between Local and Tape To Head -Motor
Phase Comparator Vert -Rate Pulses) Control Circuit Adder
i 1
Field -
Low Freq High -Freq
Pulse
Er or Error
Gen
1
Demodulated Local
Tape Sync Sync Gen
Sampling Pulse
RECORD MODE: Reference Vert From Stripped Sync of Video Signal Being Recorded
PLAYBACK MODE:
111 NONSYNCHRONOUS: Reference Vert From Local Sync Gen or Power Line
121 VERT ALIGN: Reference Vert From Comparison of Local Vert Rate to Demodulated-Signal Vert -Sync Rate
Note: This mode also involves control of capstan motor to "track' head playing vert sync to local vert sync.
131 GENLOCKED: Same as 121 but Added 'Time- Modulated" Vert Comparison to Horiz -Sync Errors
Note: This mode also involves additional error correction to head -motor control circuitry at horiz rate.
Now go to Fig. 12 -18A, which shows the basic added circuitry needed
to obtain full genlocked operation of the tape system. The demodulated
tape sync from the playback signal is phase- compared to local horizontal
sync. The error signal is split into two components: a low- frequency (or
phase) error which feeds the variable delay and a high- frequency (ve-
locity) error which adds to the headwheel -motor control circuitry.
The system incorporates a number of lock -sense buses. For example,
such a bus comes from the vertical-coincidence detector. When vertical
lock is achieved, the fully synchronous mode can take over as shown by
Fig. 12 -18B. Such circuits normally energize relays which sequence the
lock-up procedure.
Note that the primary signal for Fig. 12 -18B is the head -tachometer
pulse formed into a trapezoid. The sampling pulse changes through the
required modes of operation as shown. Step 3 is the final step, and the
tightest servo.
The simplest operation of the capstan servo occurs in the record mode.
In this mode, the capstan motor is electronically synchronized to the fre-
quency and phase of the head motor. In the playback mode, more recent
tape systems have added circuitry for faster lock -up and tight control in
synchronous operation.
Until the head motor is first locked up, the capstan must be driven
aimlessly from either a free -running multivibrator or an integration of
\
/Control
Head
-Track CT 240 -Hz
Pulse Gen
30 Hz
i
CT
Tape
Phase
Error Volts
Capstan Osc
IVCO' 240Hz
Nominal
60Hz Nominal
To CapstanMotor a4
Drive Amps Counter
the speeding -up headwheel motor. But if the head -speed (tachometer) sig-
nal is removed from a requirement of initial capstan lock -up, this condi-
tion will not exist.
The vacuum tape guide does not engage the head with the tape until
about 11/2 to 2 seconds after machine start is initiated. But the tape is
engaged with the control -track head ( which is mounted on the vacuum
guide) at all times. This signal is available immediately upon start of
the machine.
Also available, even before start -up is initiated, is the reference frame
pulse formed from local sync. This is always present at one side of a phase
comparator, as illustrated basically in Fig. 12 -19.
Consider the simplest mode of operation, the nonsynchronous playback
(neither vertical nor horizontal lock -up with local sync) . Now follow
through slowly and carefully. You know that the tape frame pulse, since
it is derived from the demodulated playback signal, cannot be immediately
available. But the recorded edit pulse (if any) is available immediately.
Again, if you are following carefully and using what you know about
timing relationships, you know that the edit pulse has the same relation-
ship (during initial recording) to the head vertical sync as the tape frame
pulse would have. So if the edit pulses are available and usable, the system
will lock up vertically as in Fig. 12 -20B. But this is not a requirement for
nonsynchronous operation.
CT'8130Hz)
The point is that if no edit pulses are available or if they are degraded,
the capstan can still lock up to the condition shown in Fig. 12 -20A. In
practice, the circuitry contains AND gates (coincidence gates) , OR gates,
and lock -sense buses to insert or remove reset -pulse gates.
The second step of lock -up is shown in Fig. 12 -20B. The trapezoid is
still derived from the ever -present reference frame pulse. The sampling
pulse is that just described for nonsynchronous operation. (The system
will already lock up vertically if edit pulses are present.)
After the head motor has been engaged by the vacuum guide, the tape
frame pulse (demodulated from the video vertical interval) becomes
available, and the synchronous mode of operation (as far as the capstan
is concerned) exists. It is now the job of the capstan to keep the head
playing back vertical sync coincident with local vertical sync. It will speed
up or slow down according to the command of the error signals in order
to maintain this tracking.
When framing ( vertical alignment) is not required, the head servo
drops the requirement for tva (tape vertical alignment) , and the variable
delay section is controlled only by the circuitry of Fig. 12 -18A. Then the
head motor, after a disturbance, locks immediately to the next horizontal
line (although a noncorresponding sync pulse is used) , and recovery is
much faster than waiting for vertical framing.
At the same time, the capstan must now follow changes in head phasing
(which occur after disturbances) to maintain the tracking set by the
control -track phase- adjustment knob (tracking control). So now the cap-
stan reference becomes that of Fig. 12 -20C. The servos are being con-
trolled by station sync, but you cannot count on being able to use special
effects, etc., since proper phasing may not exist with other local sources.
This is the Ampex horizontal mode, or the RCA Linelock mode.
are two possibilities: (1) the timing is incorrect, or (2) the trapezoid,
sawtooth, or sampling pulse is missing or low in amplitude.
Fig. 12 -21 is a representative circuit to illustrate the basic testing tech-
nique. This is applied specifically to ATC, but it will work with any of
the sampling error circuits in the tape system. Use a dual -trace scope with
one input at the Q1 collector (sample pulse) and the other input at the
trapezoid connection to the center tap of T1. (A test point is normally
provided at such connections.) Check first simply for the presence of both
pulses. If one or the other is missing, check back through that particular
path.
Voltage-
Video In
MATC
H Variale
Delay Line
Out
(
(
Proper Timing
(Dual-Trace Scope)
_FL_
From I-3 is
Trap Gen
Error
1 Det
THA Sig al Out
Back to E ror Det Amps
(See Fig. 6 -2.)
Q3
Fig. 12 -21. How to distinguish between time error and amplitude error.
Proper timing is evident if the sample pulse occurs near the center of
the trapezoidal slope. If proper timing is not obtained at this input to the
error detector, the fault will be found in the horizontal- alignment differ-
ential amplifier. (This measures the phasing between sampling pulse and
trapezoid to hold the variable delay line in the center of its range, using
a dc reference voltage against which to measure.) If timing is correct up
to the input of the error detector, you know the fault then lies in the
error detector itself or in following amplifiers. You will be able to apply
this type of troubleshooting technique to any pulse -sampling afc circuit,
of which there are many in tape systems.
Some of the latest systems do not have ATC on or ATC off available on
a switch. However, if it becomes evident that the ATC is malfunctioning,
you can do without it in an emergency. For example, on the RCA TR -70,
simply pulling the Cavec module out from its back connector will open
430 TELEVISION BROADCASTING
the ATC error -correction circuitry. If Cavec is not used, the jumper on
pins 28 and 29 of the back connector of the Cavec position may be re-
moved to delete the ATC function. (Check your particular system for a
similar way of removing ATC.) Everything (headwheel, guide servo,
playback equalizers, etc.) must be adjusted perfectly for color operation,
but a program can sometimes be saved by this emergency procedure.
Keep in mind the three fundamental headwheel (drum) servo func-
tions and the basic ways to recognize their performance:
1. There is some form of voltage -controlled oscillator (vco) which
works from the comparison (frequency discrimination) of the head
tachometer with a fixed reference. This can be termed a time -
constant control.
If you place the picture monitor on external sync and play a tape
in a nonsynchronous mode, you would adjust any time -constant con-
trol for minimum horizontal roll on the monitor. This is strictly a
frequency control at the horizontal rate.
2. Velocity loop gain contains high -frequency positional errors. The
gain should just be sufficient to get as fast a lock-up as possible, but
not so "stiff" that there is instability on start -up, splice, etc. If the
gain is too low, the response is "'rubbery" after a transient.
3. Phase loop gain contains low- frequency errors for proper damping
of the velocity loop. The headwheel (drum) has a natural tendency
to "hunt" at between 10 and 30 Hz. At too low a gain, there will
tend to be a low- frequency "oscillation" in the drum error servo
outputs.
Servo adjustments are generally made with a condition of maximum
load. This simply means that the guide is placed in about a 1 -mil "heavy"
penetration while such adjustments are made. Then normal penetration
should result in tight control.
application of boost torque to the take -up reel for an unnecessarily long
period. Fablock overcomes this by monitoring the frequency of the con-
trol -track playback signal and removing the boost as soon as the tape
Clock
US IBI
Delayed TP4
Type -D
Ref
Square 12 ms
Flip -Flop
Wave One Shot
D Set oTPB
Pulse Gen
U4
TP5
TP2 o
Tape
Frame 4.5 ms Invert OIP
Pulse One -Shot
Pulse Gen
Clock Level Change Interface --f
Q6 Q7 S rvo
U2
Co trol
a large phase error during this period, the rephasing is carried out at a
rapid rate.
Entrance into the wideband mode is controlled automatically by a cir-
cuit called the near -lock detector, which is enabled as soon as the control -
track monitoring portion of the Fablock determines that the tape has
come up to speed. This circuit compares each successive tape frame pulse
with the corresponding delayed reference square wave and decides whether
or not the timed edges of these signals are within a predetermined range
(approximatly ±6 milliseconds) of coincidence. Approximate coincidence
occurs only when the capstan servo is nearly locked, since the capstan -
servo sample pulse is derived from the tape frame pulse and the capstan
trapezoid is derived from the reference square wave (Fig. 12-20B.)
If the circuit determines that near -lock does not exist, it provides a
control signal which places the capstan servo in the wideband mode until
the next comparison is made. As soon as near -lock is detected, however,
the circuit reinserts the bandwidth -restricting components, thereby allow-
ing the capstan servo to function normally.
Action of the Fablock is not restricted to the initial lock -up period after
entering the play mode. If an asynchronous switch occurs at any time
during playback, the machine temporarily reverts to the wideband mode.
During this period, the large capacitors in the bandwidth-determining
circuits, which have been temporarily disconnected, retain a memory of
the previously determined capstan error voltage. Therefore, when near-
lock is achieved and these components are reinserted, the long -term mem-
ory voltage is added to the short-term error voltage already present in the
servo error amplifiers. As a result, the capstan motor quickly reverts to
normal speed without overshooting excessively.
The main Fablock board includes two circuits, the control -track presence
and speed detector (portion shown above the dash line in Fig. 12 -22) and
the capstan servo near -lock detector.
8ofr fr
oa _;..-t ó°a m N
tó fr
"-
O ¢S
g£ <
é ^ á
9
-
g
.-II"
t w
xI3 K1<g
g
c
:
; x '&1_
'--F---1--
.--11___7._"
B $
á
8g
á u Q S_ pS
ii
x
60- t
Oa O
o-
Z ó
00
y p
^ +S 5
Q
8 S
s
o-... ó
2f
o
_ S - -
/
Lo .. o , ---_
7-...
.
o
,,,
n Ñ 5 w. I( vn
., E
-0- ...á
E
4i:L<-:
S =
-¡(/ ^'
o
*5.
ry _
E
p R
..- II
-
3- 5-a
¢
'I"
vll
the record mode. The capstan motor will, therefore, rotate at the nominally
correct speed at this time and will not be affected by the capstan -servo
error. In addition, Q4 will remain cut off, thereby open- circuiting the
boost control bus and allowing normal application of boost to the take-up
reel motor.
The low level from the Q output of USA is also applied to the set input
of U5B in the near -lock detector ( described later) . As a result, U5B is
held in the set state at this time, thereby disabling the near -lock detector.
As soon as the control track reaches the required speed, U5A becomes
set, as previously described. The Q output therefore goes high and enables
the near -lock detector. In addition, Q5 becomes cut off, thereby enabling
the capstan servo.
The Q output of USA also goes low at the same time, but Q4 still re-
mains cut off for a delay period which is adjustable by R28. At the end
of this delay, Q4 becomes saturated and disables the boost control circuit.
The delay period is required because the slack tape, which is shared
equally by the two tension arms on the transport at the instant of starting,
is all collected at the arm nearer the take -up reel by the time the control -
track detector registers the up -to -speed condition. Therefore, to prevent
this arm from bottoming and actuating the tape-break switch, boost must
still be applied for a short additional period.
xi X
1. Withdraw the capstan error module (TR -22 and TR -70 series) or
capstan oscillator module (TR -3, TR -4, TR-50, TR -60 series).
2. Locate the C11 positive terminal (TR-22, TR-22A, TR-22B, TR-
22C) or the C17 positive terminal (TR -22D, TR -70, TR-70A,
TR -70B) or the C39 positive terminal (TR -3, TR -4, TR -50, TR-
60 series) Remove the Fablock lead from this terminal and con-
.
should occur halfway through the duration of the pulse at TP4, and
it should be so adjusted with the sample -pulse delay control (R9
in Fig. 12 -23) .
The following major pieces of test equipment are required for audio
system checks and adjustments:
442 TELEVISION BROADCASTING
Test Setup
The audio - system tests are performed with a tape speed of 15 inches
per second except where noted otherwise. All audio covers and shields
should be in their proper places. The audio monitor amplifier is con-
nected for monitoring. Fig. 12 -25 shows the test -setup connections. These
connections are changed as directed in the procedures, in order to accom-
plish the specific test and adjustment in progress.
Front Rear
TP2 12J21
Vacuum
Audio Null r Audio In Audio Signal
Tube Ai 7 Generator
Volt meter
TP3 b } o
B
Q i
i
Cue Null 12.122 /
Audio Record
Amplifier
14011
Cue Record
Amplifier
14031
..." 12124
Cue In
AC
Voltmeter
1_7
s
T P4
Audio Eraseó
f'
Erase
M011
-
Bias
Audio Cue
Preamp
14171
'X
12125
Audio Cue Playback Cue Out
Playback Distortion
Oscilloscope Amplifier
Amplifier Meter
TP5 14341
14311 ,
Cue Erase J
AC
12.123
Voltmeter
TR -70 Audio System
Audio Out
Fig. 12 -25. Test setup for RCA TR -70 audio- system checks and adjustments.
Audio Test
The audio test requires that the machine be in normal operating con-
dition, with tape in the machine. Proceed as follows:
1. Connect the audio signal generator to the AUDIO LINE A IN jack
(12J21). The generator is connected across pins 2 and 3 of the jack.
Pin 1 is ground. Then switch to Video A.
2. Set the audio signal generator for a balanced output of 600 ohms.
Terminate the generator in 600 ohms.
3. Connect the distortion meter with a 600 -ohm load resistor to the
AUDIO LINE OUT jack (12J23). Pins 2 and 3 are across the output.
Pin 1 is ground.
4. Set the LEVEL control on the record amplifier module for maximum
gain (fully clockwise) .
444 TELEVISION BROADCASTING
5. Adjust the audio signal generator for a 1000 -Hz output. Adjust the
output level to -10 dBm.
6. Turn the E-E switch (on modulator afc module 509) on. The input
signal should now be present at the output and at all audio monitor
positions.
Erase -Head Tuning and Noise -Null Adjustments
In the audio -system tests and adjustments, the head -tuning and noise -null
adjustments are accomplished as follows:
NOTE: The adjustment shaft of control R33 is located on the upper side
of module 417 and slightly right from center. The module is located on
the tape transport, just above the heads.
13. Press the AUDIO ON push button on the record panel, and make a
recording with no input signal.
14. Press the AUDIO PB push button on the audio switcher. Rotate both
the LEVEL control on audio playback module 431 and the audio
monitor SPEAKER VOLUME control to fully on (clockwise) while
listening to the simultaneous playback of the recording.
15. Adjust the AUDIO TUNE control (C31) on erase oscillator module
601 for minimum noise level as heard on the loudspeaker.
16. Adjust the AUDIO DC NULL control (R88) on the erase oscillator
module for further noise reduction. This adjustment should be
performed carefully to obtain the lowest tape noise. Rotate the
control no further than necessary in either direction of minimum
level indication to find the null point. The noise level is usually
0.04 volt with the machine stopped, or 0.05 volt with the machine
operating in the play mode over this recording. Indications are ob-
tained with the ac voltmeter.
Bias and High- Frequency Control Adjustment
Proceed as follows to adjust the bias and high- frequency controls:
Distortion Test
The distortion test and associated adjustments are accomplished as
follows:
3. Rotate the LEVEL control on audio record module 401 to its maxi-
mum clockwise position.
4. Record a one-minute section of tape at this input level; then de-
crease the attenuation in 1 -dB steps to +7 dBm while recording
each level for approximately one minute. While recording, position
the MIC switch to Audio, and announce the input level into the
microphone; then return the MIC switch to the off position.
5. Rewind this section of the tape and play it back. Set the LEVEL
control on module 431 for approximately 2.0 volts, as indicated on
the output ac voltmeter.
6. Place the distortion meter in the calibrate mode, and adjust the
distortion -meter input control for a full-scale meter indication; then
read the overall distortion.
7. Reset the input control on the distortion meter as the level changes,
according to the level announced in step 4, and read the overall dis-
tortion.
8. Determine what input level is required for the distortion to reach
3 percent (if distortion is excessive, begin recording at a lower
level).
9. Reset the attenuator to the level determined in step 8, and place
10 dB of additional attenuation in the audio generator.
10. Place the machine in the setup mode.
11. Press the AUDIO REC push button on the audio switcher.
12. Mount the audio record module (module 401) on an extender
frame, and adjust the record -level meter calibration control (R47) ,
located near the rear half of the module, to obtain a zero indication
on the VU meter at this input level (10 dB below the input for
3-percent distortion).
17. Remove the module extender frame, and replace the module in its
proper place in the machine.
18. Record a section of tape at this 0 -VU recording level, and play it
back. The distortion should not exceed 1.5 percent.
19. Press the AUDIO REC button on the audio switcher. Reduce the
attenuation of the signal generator 10 dB; then press the vu -10
DB button on the audio switcher, and note that the meter indication
decreases to 0 (± 1) VU.
1. Replace all audio shields, and record a 400 -Hz tone on a section of
tape at a 3- percent distortion level; then play it back.
2. Set the LEVEL control on the audio playback module (module 431)
for an indication of 3.2 volts (10 dBm) on the ac voltmeter.
3. Adjust the input sensitivity control on the distortion meter for 0 -dB
reference at the present level setting.
4. Rewind the tape to the starting point of the 3- percent distortion re-
cording, and terminate the audio input terminals (or apply full at-
tenuation to the audio signal generator) .
5. Operate the machine in the audio record mode over the section of
tape having the distortion recording to erase the tape with the audio
erase head only.
6. Rewind the tape to the beginning of this recording, and play the
tape back. Observe that the noise voltage level is at least 55 dB be-
low the output voltage at the 3- percent distortion level. The equiv-
alent voltages are a 0.005 -volt noise level with the 3.2 -volt setting.
7. Rewind the tape to a section where the audio is erased; then place
the machine in the play mode.
8. Operate the machine into and out of the audio record mode, and
note at what points on the counter the changes are made.
9. Play the tape back, and observe that the switching transients re-
corded are at least 40 dB below the 3- percent level, as indicated on
the distortion meter.
E -E Frequency Response
The playback loop is checked for frequency response as follows ( this
circuit may also be referred to as the back -to -back function or electrical -
to- electrical circuits) :
NOTE: Keep the output level of the audio signal generator constant.
EXERCISES
Q12 -1. Name two head faults not under your control, and describe how to
check for these faults.
Q12 -2. If you notice on a color monitor that 3 or 4 bands of 16 lines each
(indicating one of the four video heads) have excessive color beats
(moiré) resembling herringbone interference, what is the most prob-
able cause?
Q12 -3. What is the quickest way to check amplitude -vs- frequency response
of the E -E, playback, and record functions?
Q12 -4. Give the proper time base to use on the Tektronix Type 529 waveform
monitor to evaluate correctly (A) the T-pulse and (B) the 2T -pulse
on the cro graticule.
Q12 -5. What test signal most conveniently evaluates luminance-to- chromi-
nance delay inequality?
Q12 -6. What color -picture impairment results from luminance-to-chromi-
nance delay inequality?
Q12 -7. What factors other than electronic circuit problems can cause poor
servo stability and slow lock -up?
Q12 -8. What factors other than electronic circuit problems or misadjustments
can cause poor audio performance?
CHAPTER 13
Quadruplex -Tape
Cassette / Cartridge Systems
I
i
As stated earlier, the ACR -25 records in high band only. The various
record modes are: (1) onto either the A or B transport, (2) onto both the
A and B transports simultaneously from the same program source, and
(3) B-to -A or A -to -B dubs within the machine. A mechanical record lock-
out switch is available on each cassette as protection against accidental
erasure of a master tape.
In addition to the standard equipment configuration, the ACR -25 can
also be used in conjunction with an AVR -1 to time -share the time -base-
correction electronics.
Since all of the AVR -1 design features are incorporated in the ACR -25,
accessories such as auto -chroma, velocity compensator, auto -tracking, drop-
out compensator, etc., are also available on the ACR -25. Two additional
QUADRUPLEX-TAPE CASSETTE /CARTRIDGE SYSTEMS 453
accessories which contribute to the back -to -back playback capability are:
(1) control -track rewrite, which allows the operator to record a new control
track on the tape if the original one is unusable, and (2 ) automatic play-
back adjust, which allows the operator to record on the cue track, in digital
form, information that will correct chroma phase, video gain, audio gain,
and black level on a tape that was not recorded correctly. This information
is recorded only once, and every time the tape is played thereafter, it
will be correct.
Many of the controls on the ACR -25 can be located at a remote point,
and the machine can be interfaced with a station computer.
Transport Console
The transport console contains two identical vacuum -threaded tape trans-
ports; a carrousel; an internal air system; head, capstan, and carrousel motor
drive amplifiers; transport power supply; main power transformer; cassette
carriage; and transport electronics bay. The cassettes are moved from the
carrousel to the transport loading area by the cassette carriage. The tape
is then withdrawn from the cassette and threaded into the transport by the
vacuum threading process. Fig. 13 -2A illustrates the transport with front
cover open. Fig. 13 -2B presents a detailed cutaway view of the transport
with a cassette in place and the tape threaded.
In order to perform the threading operation rapidly and accurately, the
capstan, audio shield, and female guide retract in unison when the tape
threading operation begins, and move into the normal operating position,
again in unison, when the tape is threaded. When the rewind and unthread
operation occurs, the same events take place with the additional step of
high -speed rewind. Rewinding is done without head -to -tape contact. The
design of the vacuum tape transport allows for gentle, rapid, and precision
tape handling in an enclosed, controlled environment, eliminating dust and
contamination from tape surfaces.
The vacuum columns, video and audio heads, and capstan are accessible
from the front for cleaning and maintenance. The carrousel load controls
are located on a panel between the two transports. The operator has access
to any group of four cassette bins in less than one second by pushing the
appropriate button on the load control panel. Cassettes are loaded from
the right-hand side of the carrousel through a double interlocked door
(Fig. 13 -3).
Control Console
The control console contains video, audio, signal, and waveform moni-
tors and four control panels: (1) automatic, (2) manual, (3) secondary,
and (4) maintenance. The control console also contains servo-card bays,
logic -card bays, the signal system, time -base -corrector electronics, and
power supplies. Many of the printed circuit boards used in the ACR -25,
such as signal- system and time -base -correction boards, are directly inter-
454 TELEVISION BROADCASTING
changeable with boards used in the AVR -1. In addition, the Mark XX video
head is used on both machines.
Automatic Control Panel
The automatic control panel consists of a left -hand and a right-hand
portion. The left -hand portion (Fig. 13 -4A) contains a 40 -event random -
access memory which can program the 24 cassettes in any number of se-
quences and events required up to 40. This portion also contains the se-
quence display, sequence buttons, event buttons. keyboard, and keyboard
display. The right -hand portion (Fig. 13 -4B) contains the warning panel
(similar to the AVR -1) and the status display, which tells the operator the
status of both tape transports and the carrousel. This portion also contains
a number of push buttons, as follows. The AUTO button establishes that
the machine will play cassettes in the sequence as determined by the pro-
gramming of the memory. The SEMI -AUTO button establishes that the
machine will play two previously threaded and cued cassettes in sequence
and will continue to play cassettes as they are manually selected by moving
e t.
11r111
ei
the bin selector switch to another number. The LOAD CONTROL button
allows access to the carrousel for the loading of cassettes. The READY
button threads and cues the first two cassettes to be played in an automatic
sequence. The CANCEL button returns any cassettes in the tape transports
to the carrousel. The AUTO PLAY button starts the playback of an automatic
sequence. The NEXT button immediately starts the tape next in line and
ends the tape that is playing. The REMOTE button delegates selected
functions to a remote point.
control -track rewrite and automatic playback adjust, all controls on this
panel exist on the AVR -1 and serve the same purpose.
158 TELEVISION BROADCASTING
Tape t
11
Program _ Blank Tape
Program Time
6 Minutes at 15 in/s l Maximum
12 Minutes at 7 -1/2 inls
Fig. 13 -8 shows utilization of the cue track. The proposed SMPTE time
and address code (Chapter 10) has been incorporated in the ACR -25 and
is recorded on the cue track. The spare bits can be used to (A) verify pro-
gram content, (B) record automatic playback data (chroma phase, video
gain, audio gain, black level) , (C) provide a cue -up point for start of
program, (D) provide an auto -roll cue for a film projector or vtr, (E) pro-
vide an auto -roll cue for start of next cassette, or (F) provide a program
stop cue (Fig. 13 -9) .
Cassette
The cassette (Fig. 13 -10) contains a maximum of 6 minutes of program
material (at 15 in/s), plus leader and trailer. Programs may be recorded
onto a fully loaded cassette, or prerecorded tapes may be physically trans-
ferred from a reel to a cassette, thereby allowing complete interchange-
ability and avoiding further recording generations. The cassette spools
may be removed from the cassette and shipped or stored in a shipping
container.
The cassette is hinged and is easily opened for removal of the spools.
Captive record and rewind lock -out buttons are located on the cassette.
Start -of -Tape Cue Up Cue Track Auto Roll Other Au o Roll Next End -of -Tape
Marker Program Start BC Equipment Cassette Marker
systems. This section describes requirements and design solutions for the
signal -processing, headwheel- servo, and capstan -servo systems, including
their relation to conventional video tape recorders, as described by the
Ampex Corporation. In order to emphasize the reasons for particular sys-
tem parameters, a brief overview of the complete system is presented.
Overview
Most important in overall system design objectives were fully auto-
matic operation, convenient and reliable tape handling, and a cycle time
that would allow 10- second spots to be played back -to -back. To accomplish
these goals, the ACR -25 has two vacuum - threaded tape transports that are
supplied with video -tape cassettes from a shared, random -access carrousel
Leader 1 Second
[1fID111111811 Illllllllllllllllllll
11111111111111111111111111111111111111 11111111111111111111111101111111111111V. umm11nnumnnnnnnnl
Control -Track Start Digital: Program-Content Verification, Cue -Up Point Program
Begin Play to Cue Color Setup Data Program Start Start
t- Trailer --
Program
Black
7
Control Cue Automatic Roll Cue Automatic Roll Cue Program
Track Track Start Other BC Equipment Start Next Cassette Stop
storage area. Transports, carrousel, high -power drive components, and pre-
amplifiers are contained in the transport console unit. Control logic, audio
and video processing, servos, and monitoring are contained in the control
console unit. A sequence of events determined by the control logic delivers
cassettes to the transports and threads the tapes in a reliable, high -speed
manner. Once threaded, the cassette recorder is much the same as any reel -
to -reel machine, so the burden then falls upon the standard vtr components
to provide high -quality, fully synchronous video. When each tape is no
longer needed, it is rewound, unthreaded, and finally returned to the
carrousel.
A much-simplified block diagram of the ACR -25 electronics is shown
in Fig. 13 -11. The basic approach to design of the dual- transport system
QUADRUPLEX -TAPE CASSETTE /CARTRIDGE SYSTEMS 461
Transport Servo
A A
I
Sh red
Control Fig. 13 -11. Simplified block diagram
Video Out Signal
Electronics
Logic of ACR -25.
was to share as much of the signal electronics as possible, but also to pro-
vide two complete sets of servo electronics, so that tape handling could be
accomplished as easily and rapidly as possible.
Signal System
Design considerations for the signal system included minimizing the
amount of circuitry required, self- contained operation with some optional
external circuit sharing, flexible record capabilities, and fast cycle -time-
Ch I
Ch 7
Ch 4
AlB Control L yi
Equal Control
DG Control i
C 1
o 0
DG Head A
RF In From Amplifier
Head A
RF In From
Amplifier
Head B
i
Time-Base Corrected
'' I
Corrector Video Out
0-7 i
used for playback. Therefore, the servos lock to the control track, and
when the A/B switch occurs, the new horizontal-sync time may differ by
as much as -±- 1/2 line. In order to produce a fully synchronous output, the
time-base corrector must have a 64-microsecond range. This wide time-
base-correction range is an extremely important factor in simplification of
the signal-system electronics for a dual-transport vtr. A further simplifica-
tion is possible since the time-base corrector is easily shared by two vtr's
as shown in Fig. 13-14.
Flexible record capabilities are provided by two switches in the record rf
path and one switch in the input video path, as shown in Fig. 13-15. In
record-only modes, the dub switch selects input video, while the two record
Modulator Input
Video
Record Amplifie
A
Record RF
Control
Video Head Pr eamp Logic
Record Amplifier
Record Switch
Record Levels
WV, VNIsAi
External Video
From Input Filter - I
Resistive
mpedance
Matching
Resistive
Time-Base
From Time -Base Impedance
Impedance
Corrector Matching
tions more quickly and to perform some new functions. Other major com-
ponents of cycle time are determined by reel servos, the carrousel servo,
and various electromechanical components.
An important segment of cycle time is cuing. Each time a spot is to be
played, it is first threaded at a point ahead of the cue point. At the earliest
possible time, the tape is pulled forward at 150 in /s until digital informa-
tion from the cue track starts the park cycle. Fig. 13 -17 is a block diagram
of the cue servo. The basic functions of this servo are to generate a velocity
profile for parking and to determine when the tape is properly parked.
When the cue command is received, the up /down counter is preset to
488, which represents the number of capstan tachometer pulses between
Capstan
Weighted D -to -A
Converter Velocity
Reference
Parked
Capstan
Up -Down Counter
Tach Sicn
Park Preset
Up /Down
t
Capstan Capstan
Dir ction Motion
Cue
Fig. 13 -17. Cue servo in ACR -25.
QUADRUPLEX -TAPE CASSETTE /CARTRIDGE SYSTEMS 467
where the park command will be given and the desired parking point.
Contained on the cue track is a continuous playback -speed countdown of
time before the start of message (som). In both 525- and 625 -line opera-
tion, the park command is given at two seconds, seven frames before the
som. Due to the difference in tape speed, a 525 -line tape will park at 0.2
second before som, whereas a 625 -line tape will park at 0.35 second before
som. These distances correspond to the respective roll times for 525- and
625 -line operation. Receipt of the park command releases the preset input
and allows the counter to function. Digital information from the counter
drives a weighted digital -to- analog converter which produces a reference
velocity profile for the capstan servo. Three conditions must be met before
the capstan is considered to be parked and the capstan servo is put into
the stop mode. These conditions are: the tape is within ± 1 frame of the
desired parking point, capstan motion is less than 3.75 in /s, and the cap-
stan direction is such that the last increment of motion will move it closer
to the desired parking point. The total time required for the cuing process
is less than one second when the tape is threaded at a point approximately
80 inches before som.
The time required for the headwheel to come up to speed and the time
required to stop are also important segments of cycle time. Start -up time
is mainly a function of motor -drive amplifier power and may be accom-
plished during the cuing cycle. Stopping is more important. Since the tape
is rewound under reel -servo control without any head -to-tape contact, it is
necessary to stop the headwheel prior to retracting the vacuum -guide block.
Circuitry was added to the headwheel servo to provide velocity control for
deceleration down to zero speed. With the brushless dc motor of the Mark
XX head, this was accomplished simply at a low level by merely reversing
the phase of the drive commands to the six -phase motor and detecting a
stopped condition in a manner similar to capstan parking.
Automatic features common to the AVR -1 are autotracking and guide
servo, which are helpful in back -to-back operation of tapes that vary from
standard practice. Even though the servos are capable of locking up on a
badly misplaced control track, this does not occur in the 0.2 second allowed
in cycle time. For this reason, one of the servo systems is equipped with
the capability of replacing any control track with one recorded to standard
specifications. This is done once when the tape is first previewed, and
thereafter fast lock-up will always be obtained.
Fig. 13 -18 is a block diagram of the control -track rewrite method.
Basically, the capstan position loop is locked by using an advanced control -
track read head so the control track can be erased and rewritten using the
normal heads.
Module B1 generates a positional -error signal by comparing the phase of
the advanced read -head signal to the headwheel-tachometer signal. Since
an incorrect control track may be of any phase, the head -tachometer pulses
may be shifted by 0 to 360° as determined by the control -track preset, a
468 TELEVISION BROADCASTING
front -panel adjustment. Both the advanced control -track head and the erase
head are located in the video -erase stack; hence a 5 -kHz low -pass filter is
required to eliminate erase -bias frequencies from the advanced control -
track signal. The control-track filter and squarer are standard circuits as
used in the normal playback mode. Position information from module B1
is added to the normal velocity information obtained by comparing the
headwheel tachometer and the capstan tachometer.
Advanced
/Control -Track
Head
Control -Track
Ver( Raf
Erase Head
Control -Track
Preset
Tach
-Mnn- 5 Hz
Low -Pass
Filter
Motor
0 -360°
Headwheel Control -Track Control-Track
Phase
Servo Record Filter
Shift
Headwheel
r
\
/Control-Track
Head
Ref
(Head Tacht
Ph se
Comp rator -- Squarer
Module B1
Co Capstan
Motor
MDA
+ Position Loop
Velocity
Loop
Tach
Frequency
Discriminator
Tape Motion
The RCA TCR -loo makes and /or plays tv program material contained
in cartridges which hold from 2 seconds to 3 minutes of tape. The machine
holds 22 cartridges (Fig. 13 -19) The system is most useful in handling
.
QUADRUPLEX -TAPE CASSETTE /CARTRIDGE SYSTEMS 469
Control Panels
no
11611§10 ®
nono
£3000 C3 oca
Setup Adjustments
Under This Door ll
Transfer _ Transfer
Station B
Station A
Tape Decks
Courtesy RCA
short program segments such as those that occur during a station break.
Commercials and program events in cartridge form can be aired auto-
matically in sequence, with all details, including threading and video
switching between cartridges in a sequence, handled by the system.
The cartridge system utilizes a video -tape cartridge as the carrier, or
package, for each recorded segment. It has the capability of playing back
the cartridges one after another, assembling signals from each into a com-
plete, continuous program. This is accomplished by including two playback
stations ( tape decks) within the equipment. Also included are appropriate
switching and control electronics and a cartridge -changer mechanism with
a capacity of 22 cartridges. The equipment is programmed to play a se-
quence simply by placing the appropriate cartridges in the changer mech-
anism in the order in which they are to be played. It is made ready by
activating a cue -up mode, and at the appropriate time, the sequence is
initiated by supplying a single start command.
The cartridge is a molded plastic container measuring approximately 21/2
X 31/3 X 5 inches. It holds two small spools of two -inch quadruplex video
tape. It provides proper tape handling in the equipment, as well as adequate
tape protection outside of the equipment. To achieve protection, doors on
470 TELEVISION BROADCASTING
three sides of the cartridge are normally closed. They open automatically
once the cartridge has reached the playing station of the equipment. For
convenience in handling and storing, there are no protrusions on the cart-
ridge; all requirements are met within the rectangular outline.
Start of Tape Start -of -Tape Marker End -of-Tape Marker End of Tape
Maximum tape capacity is 236 feet of usable tape length plus leader and
trailer. This gives a program playing time of a little more than 3 minutes
at 15 in /s for 525 -line television systems. The recorded format on the tape
is the same as for high -band quadruplex reel -to -reel tapes. There is only
one recording standard and one tape speed (15 in /s). The recorded format
also includes several cue marks, which are necessary for proper operation
of the system. These are placed on the tape automatically when a cartridge
is recorded.
Every cartridge has two permanent mechanical (reflective) cue marks
located near the ends of the tape (Fig. 13 -20) These are used to prevent .
the tape from being completely unwound under any condition while in the
equipment. Also, the start -of -tape marker is used as the bench mark for
recording. To make ready for recording, the equipment automatically
locates the start-of -tape marker in order to determine where on the tape a
recording should begin.
Start of Tape Start -of -Tape Marker Start -of- Program Cue End -of- Program Cue End -of -Tape Marker
Active
Leader Program t
1111 _ IIi01111IIIIfl11111111111111111 _: ó
10 Sec
End of Tape
2 Sec Min 180 Sec Max
(Long -Play Cartridge)
The recorded tape format ( Fig. 13 -21) requires two additional magnetic
tone marks on the cue track of the tape to determine the exact location
and duration of the actual program segment. A start -of- program tone mark
is placed exactly two seconds before the start of the active program material.
This mark is used to determine the exact point from which the equipment
will start when it is commanded to play back. Therefore, a two -second
preroll command must be given. The cartridge recorder will start up and
provide a fully synchronized color picture in less than two seconds so that
the signal may be reliably put on the air two seconds after start -up.
An eight- second end -of- program tone mark is placed ten seconds before
the end of the program. This mark is positioned during recording by telling
the system the duration of the program, from which the system computes
the correct location for the mark. There is also a mode available to reposi-
tion this mark to correct an improperly recorded tape. The trailing edge of
the eight- second tone mark is used for the start pulse to the next cartridge
in a sequence, so that the two-second preroll is automatically provided. This
gives the concept known as sequential automation.
The magazine is the cartridge- storage device and the physical interface
between the operator and the machine. The magazine contains 22 individual
bins for storing cartridges. The bins are mounted on a continuous belt
which can be indexed on command, clockwise or counterclockwise, to get
the desired cartridge to line up with the entry portals. Indexing is accom-
plished by a photoelectric sensor, which controls the actuation of a clutch
and gear -head motor.
To transfer the cartridge into the recorder, each entrance portal is
equipped with a transfer station containing pneumatically actuated claws
and a carriage. The claws engage the cartridge, and as the carriage travels,
the cartridge is pushed through the entry portal. To retract the cartridge,
the reverse procedure is followed, and the cartridge is pulled back into the
bin.
All of the bins are continuously accessible to the operator for loading
or unloading. However, the operator needs to know which bins are being
used. For this, small V- shaped, numbered flags are used. The position of a
flag indicates the status of the bin. If the flag is tipped to the right, it physi-
cally blocks the entrance to the bin, indicating that its cartridge has been
transferred into the recorder.
In addition to the mechanical portion, the magazine contains the logic
required to perform the basic functions of indexing clockwise or counter-
clockwise and transferring in or out. A photoelectric reader provides con-
tinuous information about the belt location to the recorder control system.
The two transports contain automatic threading mechanisms (Fig. 13 -22)
in addition to the normal transport elements. To provide long tape life, the
tape path and the threader have been designed to minimize contact with
the oxide side of the tape. In the tape path, the erase heads and guiding
elements contact only the back of the tape. A urethane -coated capstan is
472 TELEVISION BROADCASTING
used to provide the necessary tape drive force without the use of a pinch
roller; the capstan contacts only the back of the tape. Therefore, the only
elements that contact the tape on the oxide side are the video, control- track,
audio, and cue record /play heads.
Besides the usual signal electronics mounted close to the magnetic
transducers, the transport contains the control logic required to thread
and unthread the tape. Photoelectric sensors have also been included to
detect fail -safe reflective foil markers placed on the back of the tape near
the start and end.
Guide Post
Cartridge
J
Fig. 13 -22. Tape threading path.
plate at the rear. The arms then move horizontally back into the equipment
to place the cartridge into playing position. Upon completion of a play (or
record) cycle, the transfer arms again engage the cartridge and move it
back into the bin.
The control system in the equipment provides all the facilities for play-
ing cartridges in a continuous sequence. This sequence is determined by the
order in which the cartridges are loaded into the belt. Fig. 13 -23 shows the
numbering of the sequential locations (bins) in the belt. It is normal to
load the belt starting with bin 1. The belt is then positioned for starting by
Thread Cue A
Station A
Cued
/d I
Changer Belt
Advance -3 Advance +3 t
Thread Cue
Station B Cued
activating the home mode, which causes the belt to move so that bin 1 is
indexed at playing station A (left) .
The machine is made ready for playing a sequence by activating the play
cue mode. This initiates the sequence of operations shown in Fig. 13 -24,
which causes both the A and B playing stations to be loaded, threaded, and
cued. At the completion of the cycle, indicators tell the operator that the
play cue function is completed and the machine is ready to accept a play
command.
Play
Video Switch Video Switch
Station A -
r_Active
1---2 Sec
Program
SS
2
1
End Cue
Sec --i
Rewind
Unthread
Thread
Cue
Sf
Play
S
Active Program
EndI
3-i
1-10 Sec--I
Rewind
Unthread
Unload 3
Load
Thread
Cue 5
Changer Belt
Un oI'. 11
I
Advance
I Load 3
Advance -5
2 -1
External Cue
Advance 2
Advance +3
Unload 2 Load 4
Play
2 Sec
End Cue
- r-
Station B Cued
Thread
h-
I
picture in less than two seconds, and the active program material com-
mences two seconds after start. The machine then plays through cartridge
QUADRUPLEX-TAPE CASSETTE /CARTRIDGE SYSTEMS 475
1 until the end cue is detected ( starting ten seconds before the end of the
program). The trailing edge of the cartridge -1 end cue is directed to start
the B playing station. (This occurs two seconds before the end of the pro-
gram on cartridge 1.) Two seconds after the start of the B station, this
station will be fully synchronized, and the program is automatically switched
from the A station to the B station. The program switch takes place during
the vertical-blanking interval and gives a transition equivalent to an
excellent electronic splice.
As soon as the program switch to the B station has taken place, the A
station automatically rewinds cartridge 1 until the start cue is passed.
Cartridge 1 is then stopped, unthreaded from the tape path, and unloaded
back into bin 1 in the belt, which has been waiting in the proper position.
The belt then moves two spaces to the left, which brings bin 3 into position
at the A station. The cartridge from this bin is then transferred into the A
station, and the thread and cue cycle is again activated. When threading
and cuing of cartridge 3 are completed, station A is ready to accept a start
trigger. Meanwhile, as soon as cartridge 3 has been transferred, the belt
moves five spaces to the right, which places bin 2 at the B station, ready to
receive cartridge 2 when its cycle has been completed.
The B station meanwhile is continuing to play cartridge 2. When the
trailing edge of the end cue from cartridge 2 is passed, a start trigger is
directed back to the A station. This station, containing cartridge 3, begins
running, and after two seconds the program switches back to A. At this
point, the B station goes through the cycle of rewind, unthread, unload,
advance belt, load, thread, and cue to ready cartridge 4. The cycle of re-
moving one cartridge and readying the next one in the same playing
station takes a little less than twenty seconds. This means that the shortest
program segment which can be sequenced continuously is twenty seconds.
However, either the first or last segment in a sequence can be shorter, since
it is not necessary to complete the cartridge change cycle at the other
playing station during either the first or last segment of a sequence. In such
cases, the cuing system is capable of handling program segment lengths
down to two seconds.
Play Control
The play control panel is shown at the upper left in Fig. 13 -26. Preparing
the TCR machine for on- the -air operation is a 3 -step procedure: pro-
gramming, previewing, and playing. (1) To program the machine, the
number of sequences and the number of cartridges in each sequence are put
into the automatic AB sequence control, and the AUTOMATIC button is
pressed. The PLAY CUE -UP button readies the first two cartridges for play.
(2) To preview, the SEQUENCE PREVIEW button is pushed; then all events
in the upcoming sequence are seen. (3) To put the machine on the air, the
PLAY button is used. (If control is from another source, the REMOTE
CONTROL button is used.)
476 TELEVISION BROADCASTING
Courtesy RCA
Fig. 13 -26. Play and transport control panels.
Transport Control
The transport control panel is shown at the upper right in Fig. 13 -26.
These controls select the transport option in which the machine is to
operate and position the cartridge magazine. Provision is also made for
cartridge reject and resetting of logic functions. The left -hand portion of
the transport control panel is a 40 -item illuminated status indicator. Color -
keyed displays of red, orange, and green indicate relative importance. As a
result, the operator is always aware of what is happening inside the
machine.
E
MESSAGE END TIME SURI -NE SHIFT EXTEOURCE
PIT-POLL
RECORD CONTROL
Courtesy RCA
Fig. 13 -27. Record control panel.
QUADRUPLEX-TAPE CASSETTE /CARTRIDGE SYSTEMS 477
Record Control
The record control panel is shown in Fig. 13 -27. Cartridge recording on
the TCR -100 is a three -step procedure: preset, cue -up, and recording. To
preset the machine, the message length is set on the MESSAGE END TIME
thumbwheels, and the external- source preroll time is set on the EXT SOURCE
PRE -ROLL register. Next, the RECORD CUE -UP button readies the cartridge
to be recorded. Then, a touch of the RECORD button starts the recording
process. Following completion of the recording, the machine automatically
rewinds and recues the cartridge for immediate replay. A touch of the
PLAY button rolls the tape for review.
System Electronics
I he system outlined in the preceding description could basically consist
of two independent automated vtr systems coupled together with a common
control system and a common cartridge changer belt. However, in such a
system, it is not necessary to provide two complete electronic systems for
the two vtr's because they are never used simultaneously. More specifically,
it is completely possible to time -share the video - signal electronics. This is
the most expensive portion of the tape-recorder electronics, so a major
economy can be realized.
The sharing of video -signal electronics in playback is accomplished by
performing the switching between the two playback stations while the
signal is still in the fm format as recorded on the tape (Fig. 13 -28A) .
Therefore, only one set of electronics- including demodulator, mono-
chrome and color time -base corrector, velocity and chroma error com-
pensator, dropout compensator, and signal -processing amplifier -is neces-
sary for the two playback stations. However, to allow for one playback
station to start while the other station is still on the air, it is necessary to
provide two complete servo systems. It is also necessary to have an fm play-
back preamplifier, an equalizer amplifier, and an fm switcher for each of
the two headwheel assemblies. With this equipment arrangement, the
system will provide instantaneous signal switching between the two tape
playback stations with excellent timing stability of both signals.
Since the video-playback signal electronics is identical to that in presently
available high -band video recorders, the ultimate economy is achieved
when the cartridge system "borrows" the signal electronics from an existing
reel -to -reel vtr (Fig. 13 -28B) . In this case, the cartridge system is a "slave"
to the reel -to-reel vtr, which becomes the "master." The master /slave ar-
rangement is accomplished by modifying the master machine to provide
an fm switch which selects between the normal fm output of the reel -to-
reel electronics and the output of the cartridge slave unit. This switch is
also arranged to operate during the vertical -blanking interval, and the servo
system of the master machine is modified so that an excellent video switch
transition is possible between the master machine and a cartridge. This
478 TELEVISION BROADCASTING
r r
He dwheet Headwheel
A sembly Assembly
_1_1 T T
4-Cha nel 4-Cha nel
Se vo
lifers Servo
Am Amplifiers
Sys em System
and witcher and Switcher
Headwheel Headwheel
Assembly Assembly
Slave Unit A B
Slave
Mono & Color Video
Headwheel Demod Timing Proc
Assembly Correction Amp
Master
Dropout
Comp
4-Channel
Se vo
Am Idlers
System
and Switch r
Courtesy RCA
Fig. 13 -29. Control electronics in servicing position.
Courtesy RCA
Fig. 13 -30. Setup adjustneant panel with door open.
Courtesy RCA
Fig. 13 -31. Rear view of TCR -100.
QUADRUPLEX -TAPE CASSETTE /CARTRIDGE SYSTEMS 481
correction devices are still in center range. Fig. 13 -30 illustrates the setup
control panel ( the narrow panel above the cartridge magazine) .
Both tape decks roll out on slides to allow routine maintenance pro-
cedures. Access to the transport decks is also available from the rear. Fig.
13 -31 is a rear view of the TCR -100.
EXERCISES
The video disc recorder, which essentially uses the same basic recording
and playback techniques as the magnetic tape system, is rapidly becoming
standard equipment in the television broadcast industry. The initial appli-
cation of this system was in the slow-motion, stop- action type of "instant
replay," and this is still one of the major functions performed by video
disc recorders. However, its use has expanded in recent years into the field
of providing automatic and practically unlimited varieties of special effects,
editing, and animation techniques for creative television. Its possibilities
in special teleproduction applications are rapidly expanding and are limited
only by the imagination of directors, producers, and artists.
We will first examine the basics of the single -disc, two -head system. Then
we will progress to the Ampex two -disc, four-head system, which is the
most commonly used video disc recorder at the time of this writing. If the
reader will study the contents of this chapter thoroughly, he should be able
to keep abreast of future developments in this field, since certain funda-
mentals serve as the foundation for all such systems.
The Ampex HS -100 is a two -disc, four -head recorder /reproducer system;
it is quite popular for a large number of special applications at the time of
this writing. We will first explore the general features of the equipment,
and then progress into more detailed analysis of circuit functions. Much of
the electronics is the same as that described for the conventional quadruplex
recorder. We will, therefore, emphasize the circuits that differ from those
already covered.
General
The Ampex HS -100 slow-motion video recorder and reproducer (Fig.
14 -2) is a transportable instant -replay television recorder /reproducer man-
VIDEO DISC RECORDING SYSTEMS 485
Control Unit
ufactured for use in studios, mobile vans, or indoor remote broadcast sites.
It is capable of recording standard NTSC color or monochrome video
signals and then immediately replaying the recorded material, either for-
ward or in reverse, at normal speed, twice normal speed, one-half normal
speed, one -fifth normal speed, or at manually controlled, continuously
variable speeds ranging from stop- action (freeze) to normal. The HS -100
also permits operator-controlled, single -frame video advance, either forward
or in reverse, whenever playback is in the freeze mode. The video output
fully complies with NTSC and FCC standards regardless of speed or direc-
tion changes. Storage capacity is 1800 television fields, corresponding to 30
seconds of video material in the normal record mode, or 60 seconds of
material in the alternate-field record mode ( i.e., recording only every other
television field).
486 TELEVISION BROADCASTING
The HS -100 is packaged in four units: the disc servo unit ( dsu) , the
electronics signal unit (esu) , the output processing unit (opu) , and the
control unit. Each unit is enclosed in a weather -resistant metal cabinet.
Carrying handles on the cabinets facilitate movement of the equipment
during transit and installation. For operation, the three larger units (dsu,
esu, and opu) can be stacked vertically with the dsu on top, the esu in the
center, and the opu on the bottom; or they can be placed side by side with
the esu in the center. All controls required to operate the HS -100 are on
the control unit. The control unit connects to the esu through a single 30-
foot multiconductor cable (200 -foot cable extensions are available on
special order) and can be placed in any convenient location. Features of
the individual units are described in the following paragraphs.
the disc servo unit, the output processing unit, and the control unit.
Cooling for the esu circuits is provided by an internal blower.
Output Processing Unit (OPU)
The output processing unit contains a standard Ampex Amtec, a stan-
dard Ampex Colortec, and an Ampex processing amplifier. These assem-
blies are mounted on slide rails in the opu cabinet for easy maintenance
access. All normally used operating controls (used for initial equipment
setup) are on the front of the individual assemblies. The cabinet is equipped
with hinged front and rear doors, a connector panel, and a blower. Electri-
cal connections to the opu are made at the connector panel at the bottom
rear of the unit.
Amtec Time -Element Compensator -The Amtec time-element (time -
error) compensator is a self- contained electronic assembly. Most of the
Amtec circuits are contained on plug -in printed wiring board assemblies
which can be easily removed when the Amtec assembly is pulled forward
on its slide -rail mountings. All operating controls for the Amtec are
mounted on the right side of the assembly front panel.
Colortec Direct Color Recovery System -The Colortec direct color re-
covery system is, like the Amtec, a functionally self -contained electronic
assembly, utilizing plug -in printed wiring boards. All operating controls
for the Colortec are mounted on the right side of its front panel.
Processing Amplifier -The processing amplifier is physically similar to
the Amtec and the Colortec in that it employs solid -state circuitry on
printed wiring board assemblies. Operating controls and test points are
located on the right side of the front panel.
Control Unit
The control unit contains all the primary controls used to operate the
HS-100. It connects to the electronics signal unit by means of a multi -
conductor cable. All controls are illuminated push buttons except for a lever
used for variable slow- motion speed control. All internal circuitry is
mounted on one printed wiring board assembly. Mounted above the con-
trol panel is an illuminated clock -type dial, calibrated from 0 to 30. A
white pointer on this dial indicates the head location relative to the 30-
second storage capacity of the system. A red pointer tracks with the white
pointer. This pointer can be stopped and then reset to its tracking position
at the option of the operator to provide a cue marking. A red on -air lamp
is provided at the top of the control unit. This lamp can be wired into the
studio master control panel if desired.
Record Process
In the HS -100, video signals are recorded on the four surfaces of two
metal discs rotating about a common vertical shaft. Recording is con-
tinuous until the operator overrides the record mode by selecting a repro-
duce or fast search mode. As long as recording continues, the latest 30
seconds (60 seconds for alternate-field recording) of recorded video is
maintained in storage, ready for instant playback. Material recorded prior
to the 30- second storage limit is progressively erased to permit recording
of new material.
Disc Rotation-The rotation speed of the two recording discs is 60
revolutions per second (3600 rpm) This speed is precisely controlled by
.
a disc -drive servo system which instantly detects and corrects drive -motor
speed variations (Fig. 14 -3) . The primary purpose of the disc -drive servo
is to lock the rotation of the discs in phase with the external reference
vertical sync. This phase lock ensures that each complete revolution of the
discs corresponds exactly to one television field, beginning and ending
during the vertical blanking period. The disc -drive servo comprises an
optical tachometer, driven by encoders on the disc drive shaft, the electronic
circuits associated with the tachometer, a velocity discriminator, a phase
comparator, the motor drive amplifiers, and the disc drive motor.
Head Transducer
To/From Head (1of 41
Preamplifier
r
Carriage
End Stop Stepping
Sensors Motor
(1 of 41
i
Logic and Disc
Composite
Swit hing Drive
Sync Input
Circuits Motor
1
Disc Motor
i Drive Control
andAmplifiers
Operator Co trol (Phase and
Commands Unit Velocity)
Fig. 14 -3. Simplified diagram of disc -drive and carriage- stepping systems.
VIDEO DISC RECORDING SYSTEMS 489
Head Movement -Each of the four head assemblies used in the HS -100
is moved by an independent stepper assembly that steps the head radially
across the surface of the disc. For purposes of identification, the heads and
their associated signal paths are referred to as head (or channel) A, head
(or channel) B, head (or channel) C, and head (or channel) D. In addition
to recording, each head also functions as a playback and erase head. During
operation, head A steps radially across the top of the top disc, head B
steps across the bottom of the top disc, head C steps across the top of the
bottom disc, and head D steps across the bottom of the bottom disc.
Each head assembly consists of a ferrite head transducer and two ferrite
pads mounted so that the head and the two pads extend perpendicularly
from the corners of a triangular platform. The head and the two pads pro-
vide a stable three -point contact with the disc surface, and they are held
against the disc by a cantilever spring which bears against the rear of the
platform. The head carriage assembly, which moves the head assembly, is
mounted on guide rails extending radially across the disc surface. As the
head carriage is moved a given distance along the rails, the head transducer
is carried across the disc in the same direction and for the same distance.
Driving power for the carriage assembly is provided by a stepping motor
controlled by logic circuits and by end-stop sensors ( which provide revers-
ing commands) The carriage assembly is coupled to the shaft of the
.
stepping motor through a pinned stainless -steel belt. The manner in which
the stepping motors move (or step) their associated carriage assemblies is
determined by the track format for the mode of operation being used.
Track Format, Normal Record Mode -Each field is recorded on the disc
as a circular track; the head is held stationary while the disc makes one
complete revolution. When head A completes recording a single field,
head B starts recording the next field. While head B is recording, head A is
being stepped to a new position. When head B has recorded its field, head
C records the next field. While head C is recording, heads A and B are
both being stepped to new positions. When head C has completed record-
ing one field, head D starts recording, head A erases the track in which it
is now positioned, and heads B and C are stepped to new positions. When
head D completes recording one field, head A starts recording in the track
it just erased, head B erases, and heads C and D step to new track locations.
In this manner each head records every fourth field, and successive fields
are recorded by heads A, B, C, and D in rotational sequence. Heads A and C
record odd -numbered fields; heads B and D record even -numbered fields.
Recording, moving, and erasing follow a definite sequence for each
head. For example, during field 1, head A records. During field 2, the
stepper assembly moves head A 0.010 inch radially across the disc ( tracks
are 0.007 inch wide, 0.010 inch center -to- center) During field 3, head A
.
which it is now positioned. At the start of field 5, the erase current to head
A is switched off, and head A is fed the fm signal output of the record
amplifier ( part of record signal electronics) The sequence for each head .
the head-carriage logic circuits detect it and correct it at the end stops
before allowing the carriages to reverse direction.
Alternate -Field Recording -In the alternate-field recording mode of
operation, head sequencing is the same as for the normal recording mode,
except that it is triggered at half the rate. The rotation speed of the discs,
the track width, and the track spacing are unchanged. The sequence is
shown in Table 14 -2. Note that only the odd -numbered fields are recorded.
During even -numbered fields, no steps are triggered, and the record
amplifier is turned off. Erase current, however, is permitted to flow during
the even-numbered field intervals.
Record Signal Path -The record signal path, shown in Fig. 14 -4, con-
tains a modulator (located in the electronics unit) and a record amplifier
and four preamplifiers (located in the disc servo unit) The modulator
.
11
E S°aÉ
o
9_
oi g1Sd
5,,
r.
E
á
óg .ti"
-
1
w á 2
.
i12%
p
,=L
0 3 w W I4o
produces line interlace of the two fields that constitute a frame. When, as
in the case of still framing, each successive field is derived from the same
recorded track, and is therefore identical to the one preceding it, inter-
lace must be restored artificially.
The phasing of the record switcher in the HS-100 is such that each
recorded field begins and ends just after the last equalizing pulse of the
-
vertical interval (point A' or B' of Fig. 14 -5) Odd fields, as recorded on
.
A A'
L.._
Fig. 14 5. Record switcher phasing.
the disc by heads A and C, begin at A' and end at B'; even fields, as
recorded by heads B and D, begin at B' and end at A'. To produce line
interlace artificially, odd fields are changed to even fields, or even fields are
changed to odd fields, by insertion of a half-line delay in the video signal
during the horizontal- scanning interval of each field (i.e., from A' to B,
or from B' to A) . The half-line delay insertion is controlled by the system
logic. By knowing what type of field is required (by examining reference
sync) and knowing what type of field is being reproduced by each head,
the logic controls the insertion of the half-line delay as required, always
removing the half -line delay during the vertical interval, A to A' and B
to B'. Accordingly, during still- framing, the half -line delay is inserted
during the horizontal- scanning interval of alternate fields.
Chroma Phase -In the NTSC color system, the frequency relationship
between the chroma subcarrier and the horizontal and vertical scanning
494 TELEVISION BROADCASTING
rates is such that the chroma phase advances 180° each line and each frame
(dot interlace) This minimizes chroma-luminance cross talk, since the
.
after; if it was bypassed before the switch, it remains bypassed after. Simi-
larly, no inversion of the chroma phase is carried out at this time.
Reverse Motion In
reverse -motion playback (Table 14 -4), the se-
quence of carriage motion and head switching is reversed from that shown
in Table 14 -3, and the carriages are made to move in the opposite direction.
Thus the fields are played back in a sequence opposite to that in which they
were recorded.
The head switching sequence, D-C -B -A, preserves the normal progression
of fields from odd to even, but loses the track -to -track phase continuity of
the chroma signal. In switching, for example, from head D to head C,
switching is from the end of one field to the beginning of the one which
preceded it in the original recording. This constitutes a 180° chroma -phase
reversal, which must be corrected by reversing the chroma phase in the
chroma inverter. Thus, in switching from track to track in the reverse -
motion direction, the half -line delay is not altered, but the chroma phase
is reversed by the chroma inverter.
For rescanning a track in slow-motion reverse, the action of the half -line
delay and chroma inverter is identical to that for rescanning in the forward
direction.
Alternate -Field Playback-Recordings made in the alternate-field mode
differ from those made in the normal mode in that only odd fields are
recorded ( the first field of each frame) . Since all fields are odd, it is neces-
sary to change the state of the half -line delay (i.e., either insert it or remove
it, as the case may be) each time the signal is switched from one track to
the next, both in the forward and reverse motion directions. In going
498 TELEVISION BROADCASTING
These signals are fed to the processing amplifier, which combines the
composite video and the time -corrected burst (added to back porch of
composite video) to provide a time -corrected composite video signal. The
processing amplifier provides two composite video outputs, one non-
composite video output, and one sync output.
Fast Search
Fast search is used to move the heads rapidly (at about four times normal
speed) from one point on the discs to another. In fast search, as in normal
operation, the heads must remain precisely in step; otherwise loss of field -
to -field continuity would result in subsequent playback. Therefore, the
sequence of motion is kept the same as in normal -speed operation. Be-
cause of the inertia of the carriage drive system moving at search speed,
it is not convenient to reverse the direction of travel of the carriages at
the inner and outer limits of travel. Therefore, a lamp and photocell ar-
rangement, located on carriage drive A, detects the approach of the heads
to the inner and outer limits and briefly slows the carriage speed to normal
while the carriage direction is being reversed.
Modulator Module
The modulator module contains circuitry to provide feedback -stabilized
amplification and dc restoration of the composite video signal before it is
applied to the fm modulator. A fast -switching voltage -controlled multivi-
brator is used to produce frequency modulation with low distortion. The
frequency-modulation process is similar to that used in video -tape recorders
in which the video information is contained in the form of sideband
energy. The carrier frequency is the same as that used in the high -band,
525 -line Ampex VR -1200 and VR -2000 video -tape recorders.
during the record mode rather than being fed to all four heads simultane-
ously. One head is always recording, one head erasing, and two heads mov-
ing to new positions.
The record amplifier receives a 0.7 -volt (peak -to -peak) fm signal from
the modulator. The signal is amplified to 2 volts peak to peak, which is a
level suitable to pass through switching circuits and provide sufficient
power to drive the recording heads to saturation of the recording medium.
The signal is separated into four output channels under control of four
gating signals.
RF Switcher Module
The rf switcher and switching logic module in J12 of the dsu receives
four rf inputs, one from each of the head preamplifiers, and five control
logic signals. The four preamplifier outputs are gated through the rf
switcher for recombination, either sequentially or on an individual-field
basis, depending on the selected reproduce mode. The module also contains
the circuits generating the gate signals for record, erase, and equalizer
switching. The logic signals are produced for each of the individual signal
channels.
The switching logic is made up of four similar circuit sections, one
section for each of the four signal channels. ( See Fig. 14 -7 for a functional
block diagram of the switching logic.) Only the section for channel A is
described. Diodes X6 and X7 form an AND gate receiving the record/
playback command and the EAc signal. The AND -gate output causes driver
Q2 to go into conduction whenever both gate inputs are high. The output
from the collector of Q2 is routed to pins 3 and C of J12. Here the signal
is split into two paths, one going to the record -amplifier channel -A record
gate, and the other going to the channel -B head -preamplifier erase -gate
input. This circuit is always inhibited when the HS -100 is not in the record
mode.
Diodes X3 and X4 form an AND gate receiving the logic EAc signal and
the inverted record /playback command. This gate, whenever both inputs
are high, causes driver Q3, Q4 to go into conduction; the channel-A
equalizer gate signal is generated at the collector of Q3. The collector output
from Q3 is connected to pins 2 and B of J12 and is supplied to the rf
equalizer board in the esu. It should be noted that this AND gate is inhibited
whenever the record mode is selected.
The record /playback command is also applied to relay driver Q14.
Selection of the record mode shifts the record /playback command level to
+3.8 volts and drives Q14 into conduction, providing a current path from
+28 volts to ground through the coil of K1. The energized relay switches
+12 volts from pin 15 to pin Y of J12 to provide the +12 -volt record
signal to the record amplifier. It also switches +28 volts from pins 22 and
Z to pins 20 and X of J12 for the head relays on the head preamplifier
assemblies.
500 TELEVISION BROADCASTING
003 NI
003 Ng
093 Pd
]Y3 Pd
093 Nd
003 Pd
N Nd
003 od
003 Pd
093 Pd
OA
31.
RF Equalizer
The rf equalizer operates on much the same principle as that of con-
ventional vtr's. Its purpose is to restore the original carrier -to- sideband
relationship of the recorded information. Two cosine equalizers are used.
One is an overall master equalizer, and the other has controls for each
individual channel.
Equalizer Gate Signal
+4V AOV PB +0.7V
0511
X3, X. )-o. Driver PB -12V-
J12 LB Ch A
Q3, Q4
2 Eq Gate
FACI
X6 X7
Driver Rec+0.9V Rec-+12V
Q2 1
Record A
1 Erase B
X8. X9
Driver
Q5, 07
LF (ChB
6 { Eq Gate
Driver
06 ( Record B
( Erase C
Driver Ch
X13, X14
Q8, Q9 D Eq
C
Gate
tCC L5 X16. X1
Driver
010 a Record
Erase D
C
018.X19 Driver
o11, 012 n Ch D
Eq Gate
Driver Record D
X21. X22
Q13 Erase A
PB+4V
15 +12V
PB +1V
Rec /PB I
Y +12V to Pin 1
of Record Amp
Command( Gate
22 +28V
Switch
20 l +28 V to Head
X ( Relays
112
Note: Voltage levels indicated are approximate values only.
PB - Playback or Reproduce It Not in Playback)
Rec Record 1 Not in Record)
Demodulator Module
The demodulator is similar to that of conventional vtr's. The function is
to limit the rf signal and detect zero crossovers of the fm carrier. Pulses are
then formed by a tunnel diode and associated switching circuitry to produce
narrow rectangular pulses, one for each carrier crossover point (therefore
VIDEO DISC RECORDING SYSTEMS 503
O OL
ov,
O w
a
fI,
Y w á 0-X= VX
0-0
-
O
NA A.H
-vV91 N
á0
tc,o.
L
aN 0d
signal switches on Q10, which in turn switches Q8 and Q9 off. The voltage
at the emitter junctions of Q8 and Q9 will then become negative 5.6 volts.
The emitter output is series connected with a control, which controls the
current through the diodes and effectively controls the rise time of the
switching waveform. This emitter output is connected to six diodes in
assemblies X1 through X3 through resistors R28 and R31. The input
signal from pin 9 also switches on Q11, which in turn switches on Q14,
causing Q15 to switch on and Q16 to switch off. The resulting positive 5.6
volts at the emitter junction of Q15 and Q16 is applied through resistors
R29 and R30 to the six other diodes in X1 through X3. The negative
voltage applied through resistors R28 and R31 causes diodes X1D, X2D,
and X3B to be forward biased. The positive voltage applied through resis-
tors R29 and R30 causes forward biasing of diodes X1A, X2C, and X3C.
Therefore, the undelayed signal from pin 4 will be gated through and
appear at coupling capacitor C22. Diodes X1B, X1C, X2A, X2B, X3A, and
X3D are back -biased and block the delayed signal from pin 7. Any spurious
signal appearing from pin 7 through X3A and X3D finds its path blocked
by X2A and X2B; furthermore, diodes X3B and X3C are conducting and
effectively ground the junctions between the X2 and X3 diodes.
When the logic -signal input at pin 9 becomes less than +3.5 volts, or
inhibitive (binary 0) , the transistors in both gate drivers reverse their
states. The signal into resistors R29 and R30 becomes negative, and the
signal into resistors R28 and R31 becomes positive. This causes a polarity
reversal of all diode biases, shutting off the path from pin 4 through the
gate and switching on the path from pin 7 through the gate. The input into
the bandpass amplifier is then from the delay line.
506 TELEVISION BROADCASTING
0 >5
o
ó
O
5
5d
O -
1 ï
r- CC
K K
r
X X
L_
i.
N
o M
CE
s VI
E o
The output from the low -pass filter ( luminance signal only) passes
through amplifier Q2. The luminance gain is factory set to 1:1 by a con-
trol in the base circuit of Q4. Chroma and luminance information are
combined by the output adder consisting of X2 and Q2. The recombined
signal is coupled through amplifier Q3, Q4, Q7 and complementary emit-
ter followers Q5 and Q6. The output from Q5 and Q6 provides a low -
impedance signal to pin 6 of J315. At this point, the signal is a standard
NTSC color signal ready for further processing by the Colortec and the
processing amplifier.
Signal CH from the logic circuitry on this board controls the chroma -
inverter function. This signal is binary 1 if the chroma inverter is required
to provide a shift of 180° in the phase of the chroma subcarrier; it is
binary 0 for any other condition. Signal CH is applied to the base of Q13.
Q13 conducts if CH is 1, and the resulting collector potential of Q13
drives Q12 into conduction. Transistor Q12 is tied between positive and
negative 12 volts. A zener diode in the emitter circuit of Q12 limits the
collector potential of Q12 to approximately +6 volts when Q12 conducts.
Transistor Q13 is held in cutoff if CH is 0, and the resulting collector
potential of Q13 holds Q12 in cutoff. Transistor Q12 is not conducting,
and the collector potential of Q12 is at approximately -6 volts. Thus, the
polarity of the gate -drive signal from Q12 changes with the logic state
of signal CH and controls the chroma -inverter circuit.
Chroma - Inverter Logic -Logic circuitry contained on the choma- inverter
circuit board controls the switch -in and switch -out of the 180° phase -
reversing network of the chroma inverter. See Fig. 14 -11 for a functional
block diagram of the chroma -inverter logic circuit. (Also refer to the
function -designators table in Section 14 -5 for the meanings of some of
the abbreviations used in the following descriptions.)
Since the HS -100 operates in the E -E mode whenever a record or fast -
search mode is selected, the output from the chroma -inverter logic is in-
hibited during the record and fast - search modes. Two signals, 8 and P4
enter the board at pins 11 and 16 of J315 and are connected to NAND gate
A4A. Either of these two signals can thus inhibit the CH output of A4A.
Three operational conditions of the chroma inverter provide for correct
chroma phase switching during (1) forward reproduce, (2) reverse re-
produce, and (3) alternate -field reproduce. Each of the three conditions
includes normal motion, slow motion, and freeze motion.
Forward Reproduce- (Refer to Figs. 14 -11 and 14 -12.) During for-
ward reproduce, the input at pin 18 of J315 is binary 1. This signal (K) is
inverted by A5B, and the resulting signal goes to NAND gate A2A, caus-
ing this NAND gate to be inhibited. The output of A2A is, therefore, bi-
nary 1 for forward motion. The normal/alternate -field record switch in
the esu is in the normal position and selects R' for application to pin 20
of J315. This signal is then applied to the PK input of JK flip -flop A3
and, through inverter A5D, to the PJ input of the same flip -flop. Pin 7
510 TELEVISION BROADCASTING
O 10 Id
i
O
á â
U
a.IP ^á Q
h1 =1 0 8 0
m
$
io
°
--. 9.
o.
O ._....
ó.
-. o°o _ °
ó 3v
co
c óP
N O
---. -.- O a G
Ñ V
Ó a0 ! p S
.
-..' ó
3
o
_
:
G
t
íJ
ÿ
U
O
"Z t_
91 .s É
p n
°oñ ._,
-ú
° ° ó
SO
-
Zi Á?.
m
5
m L"- á
OP r ti
O 6
- N G_
o+
% Ó Ó
L Lg
Q CC <J U
.. N re;
15
-
E
ó
- Oo
_
..
2
-. o°
°o
r _ 5
O d-
-
-rr
Z =Q L
The condition of the chroma inverter is changed every time that the
1/2-line delay signal (R') changes from binary 0 to binary 1. Changing
of R' from binary 1 to binary 0 has no effect on the chroma- inverter state
during the forward reproduce mode.
Reverse Reproduce- (Refer to Figs. 4 -11 and 14 -13.) The only dif-
ference between the forward and reverse reproduce modes is the signal
at input pin 18. This signal is binary 0 during reverse; therefore, it does
not inhibit NAND gate A2A, and Jc pulses can now pass through this
gate. The output from this gate is Jc', which passes through NAND gate
A2B and complements output U of flip-flop Al for every Je pulse. Signal
CH changes with every change of U and results in changing of the chroma
phase. However, during slow-motion reverse, the R' waveform must also
VIDEO DISC RECORDING SYSTEMS 513
J C Ó
V
5
L L
Q K U
Ó a0
Ó
I 5o Na, _ 5
cc Z Z
Control Functions
The control logic receives its commands from the operator through the
control -unit push buttons and generates waveforms for the control of:
VIDEO DISC RECORDING SYSTEMS 515
00
o
a
O.
m
---
o É
ó -
o
El.
O.
o
Q mU U
Logic circuit boards are located in the control unit, electronics signal
unit (esu), and disc servo unit (dsu). The control -unit logic responds
to the push- button commands by generating basic command signals (re-
cord /reproduce, forward /reverse, alternate-field /continuous, slow- motion/
normal, fast forward, fast reverse) and by generating the slow- motion
speed -control waveform. In addition, the control -unit logic board contains
electronics to drive the indicator clock and cue needle, lamp- indicator
drivers, and a 4.5 -volt power regulator.
The switching -logic module located in the esu generates the switching
waveforms for the record, erase, and playback signals. Also generated are
the control signals for the half-line delay and chroma- inverter circuits.
The clock -generator circuit, which times these waveforms, is locked to
incoming reference sync by signals generated in the sync -generator module.
Supplementary timing information is received from the slow- motion
oscillator in the control unit during slow- motion playback and from the
fast -search logic module during the fast - search mode. In the alternate -
field mode, timing information is received from the alternate -field logic
circuit located on the fast - search logic module.
The logic which controls the stepping of the head carriages is located in
the dsu on the carriage logic Al and carriage logic A2 modules.
For description and analysis purposes, the logic circuits in the control
unit, esu, and dsu are shown in block diagram form in Figs. 14 -16, 14 -17,
and 14 -25. Each block represents a functional circuit element of the system
and is identified by name. Each logic input or output is designated by a
letter which indicates its function. These function designators, plus the
other secondary functions, are listed and defined in Table 14 -5.
Push- Button Logic-Nine of the twelve push- button switches on the
control unit are interlocked with flip -flop circuits. See Fig. 14 -15 for a typi-
cal circuit arrangement. Pressing one of the nine buttons causes the asso-
ciated holding flip -flop to be set, and the active output of the flip-flop
is utilized in the control -unit logic. Depression of the push button momen-
tarily transfers the switch output from the normally closed contact to the
normally open contact, resulting in a voltage -level change. Application of
a 0 -volt level (binary 0) sets the flip-flop. The duration of the applied
0 -volt level is immaterial. The return to a positive 4.5 -volt (binary 1) has
no effect on the flip-flop state. Only a binary 0 at a reset input of the flip -
flop can cause a change of state to the reset condition. The push- button
initiated signal which sets one flip-flop may be routed to the reset inputs
of other flip -flops, causing these flip -flops to be reset. This electronic inter-
VIDEO DISC RECORDING SYSTEMS 517
Fs F function that has been modified by the fast -search logic circuitry.
G Output from clock -pulse generator. A preclock pulse coincident with Ts.
J Slow- motion clock pulse. Output from slow -motion logic.
Je Slow- motion clock pulse timed by C.
K Forward/reverse- motion logic control. Output from reverse -motion
logic.
K' Output from alternate-field switch.
L Internal connection in sync separator.
L'
M Output from carriage- reversing logic. M = when carriages move
1
P_ Output from forward /reverse push button on control unit and is the
unquantized forward /reverse command. P2 = in forward. 1
Xc
XD
Ye Output from inner carriage -limit switches on channels A, B, C, and D,
YR respectively.
Ye
YD
Z.: Output from slow- motion quantizer. A preclocked form of A, the
slow- motion control waveform.
VIDEO DISC RECORDING SYSTEMS 519
lock of the flip -flops serves to terminate one mode of operation when
another mode is selected. A change from normal reproduce to slow mo-
tion, for example, terminates the normal mode (resets the normal flip-
flop) and initiates the slow- motion mode (sets the slow-motion flip-flop) .
For the following descriptions of push- button functions, refer to the
block diagram of Fig. 14 -16.
Fast- Forward and Fast- Reverse Controls (FF and FR) -The fast -for-
ward switch (FF) and the fast -reserve switch (FR) are used without flip -
flop circuits. Both circuits are similar and therefore only the fast -forward
circuit is described. Pressing the fast- forward switch removes the normally
present ground potential from the input of an inverter and changes the
inverter input to positive 4.5 volts. Inverter output equals binary 0 1
during fast search and equals binary 1 during the absence of a fast - search
mode. The fast - forward and fast -reverse command signals, FF and are P,
routed into two paths. The first path is to the control -unit connector for
routing to the esu, and the second path is to a NAND gate. The output
from this gate is binary 1 when the freeze, fast- forward, or fast -reverse
push button is pressed. The output is inverted and fed to the reset input
of the record, forward, and reverse holding flip-flops, as well as the set
+ V1
Push- Button Switch
Function 1 3
Push -Button +4.5V
Indicator
A
101(
15
Reset Path to R57 2 470
Other Flip -Flops
Set 6 0101
012
R101
Q 500
To Logici illumination
Level
k1b Clamp
X21
13
10
Q +28V
Inverter Reset I
To Logic)
Inputs I
11 12
i ó j
ILL
u
UU
r-
0.11 0
U
input of the freeze flip -flop. Thus the termination of a fast -search mode
always leaves the system in the freeze mode.
Frame- Advance Control (FR ADV) -The frame -advance switch op-
erates without a flip -flop, and the switch output is normally at ground
potential. Pressing of the switch results in a change of its output to posi-
tive 4.5 volts; the output reverts to ground potential on release of the push
button. The output from the switch is designated as signal R1 and is
routed to the esu and to the control processing logic.
Cue Control Circuit-The cue control switch is connected to the clock
input of a JK flip -flop. Pressing the cue push button changes the posi-
tive 4.5 -volt clock input level to ground potential. Successive pressings
of the cue push button cause the flip -flop to set and reset. The flip-flop is
reset to obtain a freezing of the cue pointer, and it is set when the cue
pointer is slaved to the timer hand. Remote cue is obtained by producing
the level change from 4.5 volts to ground at a remote point and routing
this level through the normally closed contacts of the switch. Since the
flip -flop state is changed at each level change, a cue may be set at the
control unit and cancelled from the remote point, or vice versa.
Slow-Motion Oscillator-Three slow- motion switches (SM 1, SM2, and
SM3) provide two fixed and one continuously variable motion speeds.
The slow- motion circuits use a unijunction- transistor relaxation oscillator
to produce slow -motion waveform Z. A change in oscillator frequency is
obtained by controlling the resistance value in the RC section. Fixed slow-
motion speeds are provided by switching in preset resistors, and variable
slow- motion speed is provided by switching in a variable resistor oper-
ated by the speed-control lever.
The slow-motion oscillator output (Z) is connected to the control
processing logic circuitry, where it is combined with the freeze waveform
(A1) and the frame -advance waveform (R1) before being sent to the
switching logic module as the slow- motion reference waveform (A) .
Clock MDA -The clock MDA drives the timing- indicator clock in
synchronization with the carriage stepping. The input signals are
(from the esu), which goes to binary 0 each time head D is recording or
reproducing, and signal K, the quantized forward /reverse command (from
the dsu) , which indicates the direction of clock motion.
522 TELEVISION BROADCASTING
Control Unit
Sync Input P4 P3 A W FF Fe AF
11111.111111j
Servo Ref SRD To Servo
Delay
Sync
Separator
Horiz Ref
To Amtec & Colortec
Delay S YD
T Pulse
Delay
TSD
WS
I [
E -E
0 Command
yy Fas
Sear h Fs Clock Slav Motion
FF Logic Generator
G
Logic
Bi
Y.-.
P25
DG
3
Delay Line Slav Motion
4 Alternate Logic Quantizer
Field
Logic
f As
To Record and
BG BG
Playback Switchers
All Field I
tttt
_
Normaall't 4.5
AF
DG EAC EBC ECCEDC
? Gnd
Head
EaG EAK Head
E EBG Retiming
Logic bG
ECK Logic
EcG
EdG EDG
L
As Ra K'
PG
Chroma
Inverter
i
¡ 7
K P2S xaa aa EAG I EBGÍ ECCI EDG EAK ECK
DisdServo Unit
pulse (F) ; and inhibition of the half -line delay command. Fig. 14 -19 is
the clock generator, the slow- motion quantizer, the slow-motion logic,
the head logic, the head retiming logic, and the half -line delay logic. Each
of these functions is described below; timing diagrams show the rela-
tionships of the waveforms.
Clock Generator -The clock generator is supplied with input signals
Ts and Fs ( Fig. 14 -20) Both input pulses originate at the fast -search
.
logic board. From the two input pulses, the clock generator produces three
basic timing pulses, which are required to synchronize the switching of
the system logic. The preclock pulse (G) , the system clock pulse (C) ,
and the timing pulse (Bo) are the clock -generator outputs.
The preclock pulse (G) is derived from the leading edge of the Ts
TIN
TS
(Normal)
J
TS
(Fast Search)
BG
waveform and is coincident with this leading edge. Logic circuits are
preset by the G- pulse, which occurs prior to system clock pulse C and
therefore assures that switching transients subside before system switch-
ing takes place. A further consideration for this arrangement is the fact
that the logic components used in the circuitry possess inherent and finite
transfer delays, which could prevent the correct completion of switching
sequences during the presence of a system clock pulse.
The Ts -pulse is applied to an inverter. After differentiation, the leading
edge of the Ts waveform produces a negative -going pulse, which is used
to trigger a 20- microsecond one -shot multivibrator. The inverted output
represents pulse G and is routed to the slow- motion quantizer and the
slow- motion logic.
VIDEO DISC RECORDING SYSTEMS
I nputs
Fs I
TS
Output s
Ts
6G
The system clock pulse (C) is derived from the trailing edge of the
Ts- pulse. Its timing is such that the leading edge of the C -pulse is coinci-
dent with the trailing edge of the Ts- pulse. The timing of the C -pulse
therefore is such that it occurs approximately 600 microseconds ( the dura-
tion of the Ts- pulse) after the G- pulse. The C- pulses are used to time the
signal- switching and carriage -drive pulses.
The BG -pulse is a divided -by -two version of preclock pulse G and is
phased by field- identification pulse FS. Phasing of BG by means of Fs
causes even fields to be recorded by heads A and C and odd fields to be
recorded by heads B and D. This condition is valid for the normal record
mode only, since the alternate -field record mode does not allow recording
of even fields. The BG waveform is a binary 1 during even fields and a
binary 0 during odd fields.
Slow -Motion Quantizer -(See timing diagram, Fig. 14 -21.) The slow -
motion quantizer takes slow- motion reference waveform A (generated in
the control unit and routed through the alternate-field logic as As) and
converts it to ZG. Output ZG is a rectangular wave, each zero crossing of
which is coincident with a G- pulse. The number of zero crossings of ZG
I n put
11111111111111111111111111111111
u ;tput
Nor
SiOw 'aOtÍP.
motion by circuits on the carriage logic board and re -enter the switching
logic board as waveforms EAK and ECK.
Head Retiming Logic -(See timing diagram, Fig. 14 -24.) The head
retiming logic reclocks all the head pulses against waveform C prior to
the application of these pulses to the head switcher. The four head -pulse
inputs and their complements are applied to preset inputs of JK flip-flops.
The complements of the input pulses are obtained from inverters made up
of NAND gates. All four flip-flops are clocked by waveform C and there-
fore produce head -pulse output changes coincident with the leading edge
of clock pulse C. The resulting output pulses-EAC, EBC, Ecc, and Enc-
ase routed through the harnesses to the rf switcher and switching logic
board in the dsu. Pretimed waveform EDG is also passed through an in-
verter. Thus EDG is routed to the control unit for synchronization of the
indicator clock.
Half-Line Delay Logic- During playback in any mode but normal-
speed forward, an odd field may be gated through the reproduce electronics
at the time an even field is required; or, conversely, an even field may be
gated on when an odd field is required. Since this would destroy the pic-
ture interlace, it is necessary to convert odd fields to even fields, and vice
versa. Conversion of one to the other is accomplished by a half-line delay
of the video signal during the horizontal- scanning interval (but not the
vertical- synchronizing interval) of each field requiring correction.
The half-line delay logic generates waveform R', which controls the
half -line delay. The delay is switched out when R' = 1 and switched in
when R' = 0. To do this, the logic determines (A) whether an odd or even
field is required and (B) whether an odd or even field is being reproduced.
If a correction is required, R' becomes 0, and for no correction R' is 1.
Also, during the vertical- synchronizing internal, R' always is held to the
value 1. Waveform BG determines what type of field is required, since BG
is backed up to reference sync and always has the value 1 during
even
fields and the value 0 during odd fields.
J
The DG pulse determines what type of field is actually being reproduced.
In normal speed, DG = BG, so odd fields are always reproduced when odd
4v
(
OV- -L
EBOI
(
V
ov I I
}
4v-
OV
fields are required, and even fields are always reproduced when even fields
are required. The half-line delay, therefore, remains switched out. In slow
motion, odd fields are reproduced when DG = 0, and even fields are re-
produced when DG = 1.
Waveforms BG and DG are connected to a NAND gate, and BG and D,
are connected to another NAND gate. The outputs are tied together and
form an exclusive OR function labelled R'. This function follows the
Boolean equation:
R'=BG DG+DG' BG
or,
R' = BG DG + BG DG
In the alternate -field mode, the action of the chroma inverter is the
same for forward and reverse motion. The chroma phase must be reversed
each time the half -line delay is inserted and each time the heads step, unless
the two occur at the same time. The chroma- inverter control waveform is
designated Cg. The chroma- inverter logic has been explained previously.
Disc Servo Unit Logic
The disc servo unit logic is located on the carriage logic Al and car-
riage logic A2 modules, which are located in the dsu (Fig. 14 -25) . The
purpose of the logic is to provide the head -stepping command pulses
(F- pulses) to the carriage -motor drive amplifiers. This logic maintains
proper head synchronization for all motion speeds, forward and reverse
directions, and inward and outward travel of the heads across the disc. The
arrival of the head carriage at the outermost or innermost tracks of the
discs is detected by means of photocell end -stop switches, which control
the reversal of head-carriage motion at these points.
Electronics Unit
lEAK G
ECK 0 IC
EAG
EBG
Reverse -Motion
ECC Carriage
Logic EEG Logic
IA21
FAG IA21
K EAKFCK ECG
Carriage Retiming
Carriage Control F0
Logic
Logic F
IA11
(Al) c
Fd
Fa Fbc Fcc Fdc
A l A
Jc X N
EBG
r M
X0 YdX
Xpa Xc Xd Ya Yb Yc
X X0
Oco
XC Xc Carriage
Carriage
Error ü0 Carriage
Reversing
Logic XD End Stops Xd Correction rd; MDA
YA Ya
IA21 Logic rTa
YB Yb (All rt-ci
Yc
YE
YD Yd
Fdcì
erasing (in preparation for recording the next field) , and two are step-
ping to the adjacent track during each field interval. For example, during
field 1 ( when EA = 1) , head A is recording, head B is erasing, and heads
C and D are stepping. Similarly, during field 2, E1 causes head B to record,
head C to erase, and heads D and A to step. Then during each cycle of
four fields, each head erases, records its field, and steps two tracks before
repeating the cycle. These pairs of stepping pulses, the F pulses, are gen-
erated in the block labelled "carriage logic" in Fig. 14 -25.
The motion of the heads across the discs is shown in Fig. 14 -26 for
normal -speed playback. The fields are displayed horizontally, the track
number vertically. Head motion is identical to that described above for
the record mode. During field 1, head A is reproducing on track 6. During
field 2, head B reproduces on its track 6, while head A moves one track to
track 5. During field 3, head C reproduces its track 6, head B moves to
track 5, and head A moves to track 4. The heads move toward track 1 in
this manner, each reproducing a field in turn and then moving twice to
the next even-numbered track. After reproducing field 9, head A steps
once, then reaches its end stop. Further motion of head A in that direction
is inhibited by the carriage error -correction logic. By the end of field 13,
all heads have reached their end stops at track 1. At this point, the direc-
tion of head -carriage motion is reversed, and head A begins moving away
from the end stop, this time reproducing the odd -numbered tracks, fol-
lowed by heads B, C, and D in the manner described above. When the
other set of end stops is reached, a similar carriage reversal takes place,
and the heads once again reproduce the even -numbered tracks.
If, instead of forward motion, the operator selects reverse -motion play-
back, the recorded tracks must be played back in the sequence opposite to
that in which they were recorded. Such a reversal of motion is shown at
the end of field 21 (Fig. 14 -26) The direction of carriage motion is re-
.
versed at this point, and the E pulses are made to reverse their sequence
(Ec- EB- EA -ED, instead of EA- EB- Ec -ED). This reversal of E -pulse sequence
is accomplished by interchanging the EA and Ec pulses. Upon selection of
534 TELEVISION BROADCASTING
It can be seen from these equations that each carriage will step twice
during each series of four E pulses, satisfying the stepping requirements
for forward -motion operation. Further processing of these signals is re-
quired to produce carriage- stepping pulses which satisfy the requirements
for carriage movement during reverse motion and while negotiating the
end stops.
Carriage -Reversing Logic-The carriage- reversing logic generates the
carriage- reversing command (M) which governs the direction of motion
of the carriages. Carriage motion may be inward or outward for forward
or reverse video motion. Inward motion is defined as carriage travel from
the periphery toward the hub of the discs; outward motion is the oppo-
site. The arrival of a head at the outer track is detected by photocell X (at
the inner track by photocell Y) for each disc surface. The carriage-
reversing command (M) is a function of two signal sources, the motion -
reverse command (K) and the end-stop switch signals.
The reverse- motion logic circuitry generates quantized reverse -motion
command K and produces the reversal of the channel switching and
stepping sequence (A-B -C -D changed to C- B -A -D) during reverse mo-
tion. This is accomplished when K = 1 by interchanging the EA pulse
with the Ec pulse, and the FA pulse with the Fc pulse.
Reverse- Motion Command K- Command K is equal to the P2s com-
mand quantized by head signal EBG and preclock pulse G. Command K is
also inhibited from changing while the carriages are at their end-stop
switches, as motion-reverse switching coincident with carriage end -stop
reverse could cause errors in the carriage reversing process.
Head Switching Sequence in Reverse Motion -Reverse video motion
is obtained by reproducing the video in a sequence opposite to the
record
sequence. Substitution of the channel -A timing signals for the channel -C
signals and vice versa converts the sequence from A -B -C -D to C -B -A -D
for reverse video motion. The head pulses for channels B and D are not
affected by reverse motion, but the pulses for channels A and C are.
Carriage Stepping Sequence in Reverse Motion -The reasons for ex-
changing the channel -A and channel -C pulses described in the preceding
paragraph for the head logic apply also for the carriage pulses. Output
signals FAK and F(.K are derived in a manner similar to that used for the
head logic, except that carriage- motion pulses FAG and FVG are used in-
stead of head pulses EAG and ECG. Four gates are contained in assembly
A2, which is used in this circuit. Pairs of gates have their outputs tied
together, performing the function of an OR gate. The Boolean equations
for each of the output signals are as follows:
FAK = (FAG ' K) +(FcG'K)
FUK= (FcGK) +( FAG g)
where,
FAG = Esc + ECG
VIDEO DISC RECORDING SYSTEMS 537
and
Fco = EDG + EAG
Fe i
F D'
E I
Time
Notes: E End Stop Activated
I Stepping Inhibited
See Fig. 14 -28. When all heads reach their end stops and the direction
of carriage motion is reversed, each head, in the sequence D- C -B -A, must
take only one step away from the end stops before resuming normal 2 -track
stepping. Head D, the first to leave, automatically moves only one track
because carriage reversal takes place after what otherwise would have
been its first step.
Carriage Retiming Logic -The carriage retiming logic changes the
timing of the carriage signals from preclock time G to system clock time
C, then samples them with the J pulse to produce the short -duration car-
538 TELEVISION BROADCASTING
V
-
FA
Time
Notes. E ° End Stop Activated
I = Stepping Inhibited
EAK
FBG
ECK
EDG
k 11111111111111 I I I I I I I
FA'
FAG' I I I I I I I I I I
F1'
F BC' I I I I I I I I I I I
FC'
FCC' I I I I I I I I I
FDC' I I I I I I
Time
Fig. 14 -29. Head and carriage timing waveforms of carriage -retiming logic.
VIDEO DISC RECORDING SYSTEMS 539
riage pulses FA(', FRC', FCC', and Fix,' It employs binary devices and asso-
ciated NAND gates and inverters. The four channels operate in a similar
manner. See Fig. 14 -29 for waveforms of the retiming logic.
Carriage Error- Correction Logic -The carriage error- correction logic
serves to inhibit the further motion of each carriage as it reaches an end
stop. When all four heads reach their respective end stops and the direc-
tion of carriage motion is reversed by the carriage reversing logic, the
carriages are allowed to step away from their end stops in correct sequence
as determined by the carriage control logic. Output pulses Fac,,, Fb,.,,, Fee.,
and Fac,, control the outward movement of the carriages. Pulses Fad, Fbei,
Frei, and Fdei control the inward movement of the carriages.
Features
Because of the instant review feature of the HS -200 teleproduction sys-
tem, animations or other edited program material are available for review
in full color immediately after recording. When the material is reviewed
in its finished form and corrections are necessary, all or any part of the
material (even to a single frame) may be replaced. The preprogramming
feature of the HS -200 system permits automatic editing of complete pro-
grams, such as commercials and animation sequences.
System Console -The console contains all the primary operating con-
trols for the system. The primary controls and indicators are arranged by
function on six panels: record panel, memory panel, slow- motion panel,
dissolver panel, numerical -display panel, and patch panel. In addition, a
picture- monitor panel and a waveform -monitor panel are provided.
Variable Frame Increment (Animaation) -The HS-100 has a total ca-
pacity of 900 frames, which may be recorded in one operation or in a
selected number of frame increments between one and 900. By this means,
540 TELEVISION BROADCASTING
in the required order at the time of editing. The process of assembly from
prerecorded material involves locating the desired sequence and attaching
it to the preceding one either physically, photographically, or electroni-
cally. This is actually the only way the operation can be done with film
with any degree of precision or artistic flexibility. With either tape or disc,
however, assembly and recording can be accomplished concurrently by
producing the scenes in order and recording each one to its desired tim-
ing and making the desired transitions (Chapter 10) The HS -200 adds
.
an order of magnitude to the ease, speed, and precision with which this
can be done, and progress may be reviewed as the operation is carried out.
Insertion-Insertion of new material into previously recorded material
is a common requirement. Accurate placement of the insertion with regard
to preceding and following material requires the ability to determine the
"in" and "out" edit points of both the new material and the material into
which it is to be inserted (Chapter 10) The HS -200 offers two possibili-
.
ties for carrying out such an operation: (1) The material to be inserted is
recorded on the HS -200 and cued accurately to the desired "in" point and
then electronically edited into the prerecorded sequence on tape from
cues previously placed and precisely adjusted. (2) The original material
is recorded on the HS -200, and the new material is inserted at the desired
points from any other source, once again with one-frame accuracy of
placement.
Cutting -The cut (an instantaneous change of scene) is the simplest
of video transitions both physically and subjectively; timing is the only
variable. With film, proper timing is accomplished by selecting the frame
at which the cut is desired. The HS -200 offers this facility by allowing
examination of material at a frame -by -frame rate, and it also allows realis-
tic preview of the cut before it is accomplished.
Dissolving -The dissolve is a more complex transition, requiring the
fading out of the first picture simultaneously with the fading in of the
second one. Timing again is the critical point, with respect not only to
when the dissolve occurs, but also to how long it will last. With film, this
is an operation that is done in the optical printer with precise control of
the number of frames involved, and therefore must be entrusted to the
lab technician. With the HS -200 and an accompanying video source, dis-
solves may be achieved with the same accuracy and automatically con-
trolled as to the length. Again, the major advantage offered by this system
is the ability to preview or rehearse the effect with the possibility of
making desired adjustments before actually performing the final transition.
Fading -The fade involves simply fading from picture to black or
vice versa. It is easily accomplished in a number of ways. It can be done
at the time of recording or during the editing process. The advantage
offered by the HS -200 in this operation is accuracy of timing.
A -B Roll -The A -B roll technique, commonly used in film production,
involves dividing the production into "odd" and "even" scenes. The odd
VIDEO DISC RECORDING SYSTEMS 543
scenes are placed in sequence on one video source and the even ones on
another. With proper timing between the two sources, odd and even
scenes are then combined into the desired final form. With the precision
of control available, this is easily accomplished with two HS-200's or by
one HS-200 and a second video source.
Keying -In this process, one picture signal is inserted into another, e.g.,
a title into a background scene. In film, this is a laborious process requir-
ing time and expense. In television, it is achieved electronically by com-
bining two video signals. The main use of the HS -200 in this process is as
the source of the signal to be keyed. Because of the frame -at -a -time play-
back capability of the HS -200, material to be keyed may be recorded on
succeeding frames and keyed into the background as desired and changed
on cue. It is possible to record animated titles or products on the system
and key them into other material with full control of movement and
timing.
Matting -This process is the same as keying as far as the HS-200 is
concerned.
Superimposure- Superimposure is a process that combines video ma-
terial from two sources. The end result of superimposure is similar to that
of keying or matting, although the effect is not as crisp. The HS -200 acts
as one of the sources, probably for printed material.
Time Lapse-This technique, which might also be called "time com-
pression,' allows the reproduction of lengthy events to be condensed as
much as desired. The blooming of a flower, for instance, can be repro-
duced in a period of five seconds. This effect is accomplished by taking
picture frames at selected intervals during the event being recorded, rather
than recording it continuously. When the reproduction is played back con-
tinuously, time compression is achieved. In other words, if one frame is
recorded every five seconds during a 15- minute process, the final record-
ing will have a total of 180 frames. If these frames are played back at the
normal rate (30 frames per second for television or 24 frames per second
for film), the whole event can be viewed in six seconds. This is done on
film by exposing one frame (either manually or by a timer) at every
selected interval. The same technique is possible on magnetic recordings
with the HS-200. Two advantages accrue: (1) the reproduction is imme-
diately reviewable, and (2) playback is not locked to any particular frame
rate (as it is in film) , allowing the length of the event to be manipulated
to fit any time segment.
Simulated Time Lapse-This interesting technique, which produces the
effect of time lapse, is achieved by combining the backward- motion and
single -frame recording capabilities of the HS -200. For example, a potted
plant may be placed on camera and a series of frames recorded with a
little of the plant snipped off between each "take" until the plant has been
cut entirely away. If this sequence is played in reverse, the plant seems to
grow before the camera. This technique can be applied to a number of
546 TELEVISION BROADCASTING
given instant. In freeze operation, for example, the frame counter is sta-
tionary. When the HS -200 is either turned off or placed in the standby
mode, the system may develop a counting error. To resynchronize the
frame counter after turn -on, the controls are set for normal forward op-
eration, and all four heads are allowed to move to the outer edges of the
disc.
There are eight stores accessible through the memory panel. The num-
ber in the store being accessed at any time is shown on the store display.
However, the particular store being accessed on the memory panel is indi-
cated by an illuminated store push button.
MODE
STANDBY RECORD PLAYBACK
ALT FRAME FIELD
OPERATE
FRAME FIELD
SYNC
LIGHT
O
RECORD RECORD START STOP
TIMER
READY
R T S S
the HS -100 (CB -1, rear connector panel of esu) and HS -200 console (in-
side lower left front panel) must be switched off.
The OPERATE /STANDBY push button controls the power to only the disc
motor and hour meter, and has no effect on the remaining circuits, cotn-
puter logic, or information in stores. Setting this control to Standby will
save unnecessary wear on the heads and discs.
The green SYNC indicator lamp is in parallel with the lamp of the disc
servo unit. When the lamp is fully on (not flickering), in indicates to the
operator that the disc servo is in sync. Neither a correct recording nor a
correct reproduction can be achieved unless this lamp is on.
The RECORD READY control takes the system out of the reproduce mode
and prepares it for instant recording in response to a record command.
When the RECORD READY push button is pressed, the SYNC lamp will flash
momentarily, then glow steadily, indicating that the system is ready to
record. NOTE: The system will not switch to record ready from the re-
verse reproduce condition; it must first be placed in forward.
The start, stop, and record timer functions are best explained by de-
scribing manual, automatic, and programmed -sequence recording opera-
tions. To make a manual recording. the record timer must be off and the
system must be in record ready. With the green SYNC light steady and the
START push button pressed, the system will begin recording on the next
15 -Hz color editing pulse. The system will stop recording when the STOP
button is pressed. To make each successive recording, press the START
button and then the STOP button.
To record a preset number of frames automatically, the record timer
must be on and the system must be in record ready. When the START
button is pressed, with TIMER STORE HOLD off and SEQUENCE STORES
(Fig. 14 -31) off, the HS-200 will wait for the next 15 -Hz color editing
pulse, after which it will record for the number of frames displayed on
the store display. The same number of frames is recorded each time the
START button is pressed.
Up to eight different recording lengths can be programmed in sequence
instead of one at a time. The length of the first recording in frames is put
in store 1. The length of the second recording in frames is put in store 2,
and so on up to store 8. When this is completed, the RECORD TIMER, RE-
CORD READY, and SEQUENCE STORES push buttons are all set to the on
position, and the TIMER STORE HOLD button is off.
The first time the START button is pressed, it will cause the number of
frames in store 1 to be recorded. The second time it is pressed, it will
cause the number of frames in store 2 to be recorded, and so on. As al-
ways, the manual operation of the START push button can be replaced by
a cue from a vtr (by feeding a tone -burst cue, for example, and patching
from Remote VTR A to Start Record) .
Pressing the TIMER STORE HOLD push button to the on position stores
the number shown on the store display in a ninth store called the timer
548 TELEVISION BROADCASTING
store. The number will be stored there until the TIMER STORE HOLD button
is pressed to off. If the system is in automatic record, it will record for
the number of frames in this ninth store. With a pencil, the operator
should write the number stored in the space provided above the TIMER
STORE HOLD push button. After the number is stored, the display may be
changed without affecting the number stores. The use of the timer store
hold feature is recommended, instead of methods described which do not
use this feature, for making an automatic recording. If the wrong store
is accessed by mistake when the TIMER STORE HOLD push button is off,
2 3 4 5 6 7
STORE STORE STORE STORE STORE STORE STORE STORE
ENTER
ENTER
CUE
C E
17 8 9
5 6
2 3 SEQUENCE
.11
STORES
o 0
R
SEARCH
START
PROGRAM
S P
STORING FUNCTION
OPERATIONAL FUNCTIONS
System Operation
We will go through a brief description of operating conditions and the
philosophy associated with system control. Detailed step-by -step procedures
are provided in the instruction book for the HS -200 system. Obviously,
space does not permit such details in this text, and only fundamentals im-
portant to a general understanding of operations procedures are presented.
550 TELEVISION BROADCASTING
will occur from the 18th to the 24th second of the total time, the scene
can be programmed on the HS-200 to start recording at frame number 560
and stop 180 frames (6 seconds) later, at frame number 740. If it is being
recorded "live," the action is cued, and the HS-200 is started manually.
Only the number of frames desired will be recorded, and the next scene
can then be recorded in a similar manner by again adjusting the number of
frames to be recorded. If the material is coming from a video -tape re-
corder, a cue may be placed at the "in" point on the tape, and this will start
the HS -200 in record automatically. In this way, precise timing adjust-
ments of scenes and cuts are possible.
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
300 Frames - -f- 240 Frames 180 Frames -1 -180 Frames
Total Sequence
Now, if it is desired to go back and record scene II, just determine the
length of scene II (in frames) subtract that from the starting point of
,
scene III, use the resulting frame number as the starting point of scene II,
and program the HS -200 to record the number of frames in scene II.
An entire sequence would be recorded as follows:
A. Cue the HS -200 when the frame counter displays 560; then program
the system to record 180 frames. This will record scene III.
B. Without changing the frame number on the HS -200 or the number
of frames to record, record scene IV, which is exactly the same length
as scene III and follows directly after it.
C. Next, to record scene II, which is 8 seconds (240 frames) in length,
cue the HS -200 at frame count 320, which is exactly 240 frames
before the beginning of scene III. Set the record timer for 240
frames and record.
D. Finally, record scene I in an identical manner, cuing the HS -200 at
frame readout 20 and setting the record timer for 300 frames (10
seconds) , the desired length of scene I.
VIDEO DISC RECORDING SYSTEMS 551
FROM T TO
Oo 3 3 Qo ©
Slow Motion
Slow Motion
2
3
00 o o
00 4 4 O Qo Freeze 00 o o
O 5 5O O Single Frame
O o o
©6 6 0 ® Forward
07 70 Oo Reverse O cap®
BO
OB
00 o o
Seq Patch
Address Panel
O O Oo Dissolve to Chan A O Key in
Input On /Off © Dissolve to Chan B 00 Key Out
Ground Patch
© Toggle Dissolve 00 Toggle Key
Next, connect the HS-200 playback output to the video input of the vtr,
and play back the tape from the beginning. The "in" cue will start the
HS-200 in playback and permit the output to be viewed on the monitor
of the vtr. The Editec "out" cue will switch the monitor back to vtr output,
providing continuous rehearsal of the entire sequence. If it is satisfactory,
recue the vtr and HS -200 as before, and start the vtr in the Editec record
mode. On this pass, the insert will be recorded exactly as seen in the
rehearsal.
Transitions -The HS -200 can be used to make a cut from one scene
to another or to assist in making a cut. If the total time of the two scenes is
less than 30 seconds, both can be recorded on the disc with precise timing.
The completed sequence of two scenes may then be transferred back to
tape by means of Editec, allowing precise placement of the finished
product.
Cuts may also be achieved on A -B rolls, using two vtr's and an HS -200.
In this case, either the A or B scene may be recorded on the HS -200, and
the other recorded on a vtr. Then, with both signals fed through the auto-
matic dissolver to a second vtr and with playback timings preprogrammed
(by means of Editec and HS -200 cues), the entire sequence can be recorded
in one automatic operation.
Dissolves are accomplished in much the same way at cuts, using the
HS-200 as the source of one signal and a vtr or live camera as the other.
The advantages are that the automatic dissolver in the HS -200 allows
complete control of both placement and length of the dissolve and allows
this operation to be carried out without tying up studio switching equip-
ment. Because the dissolver will dissolve in either direction (from or to the
material on the disc), it is possible to include a number of dissolves on
an A -B roll.
Since a fade is really one -half a dissolve (or a dissolve to or from black) ,
this effect is also readily available with the HS-200. Black is used as the
alternate source to the HS -200 signal, and the fade is programmed for the
rate and placement desired.
Although the patch panel (Fig. 14 -33) can be programmed to control
every function of the HS-200 automatically, either separately or in com-
bination, it is not a complex operation. System functions are enabled
through the patch panel by means of cue pulses, which are directed by
patch cords to the proper inputs. These cue pulses may come from an
outside source such as another HS -200, a video -tape recorder, a manual
control at a remote location, etc., or from within the HS-200 by means of
the memory stores. (It may help to think of these pulses as an extension
of the operator's finger which will push the proper buttons at a preselected
time.) In order to program the system, the cue pulses must be routed from
the source on the FROM side of the patch panel to the proper inputs on the
TO side of the patch panel. In other words, during preprogramming, if the
operation requires selection of two functions together, it also requires a cue
VIDEO DISC RECORDING SYSTEMS 553
tween successive recordings. The speed of the animated action is, of r ourse,
controlled by the speed of playback, either forward or reverse. '
t
VIDEO DISC RECORDING SYSTEMS 555
EXERCISES
Q14 -1. In the HS -100 system, how much video material is available on the
disc in the (A) normal record mode and (B) alternate -field record
mode?
Q14 -2. How many disc surfaces are involved in the HS -100 system?
Q14 -3. If a recording made on the video-disc system runs longer than the
maximum capacity of the disc, what happens?
Q14 -4. What is the speed of rotation of the video disc?
Q14 -5. How many lines of video information are recorded during one
revolution of the video disc?
Q14 -6. Does the video head move while it is recording on the video disc?
Q14 -7. What is the major difference between the feeding of record currents
to video -disc heads and the feeding of record currents to video-tape
heads?
Q14 -8. How wide is the recorded track on the video disc?
Q14-9. What is the (A) center -to- center spacing of recorded tracks and
(B) guard band between tracks on the video disc?
Q14 -10. Is speed -up motion available as well as slow motion?
N
APPENDIX A
Table of Prefixes
Power of Ten
Prefix Symbol Definition Multiplier
atto a One millionth of one millionth 10 -18
of one millionth
pico* p One millionth of one millionth 10 -72
micro- µµ One millionth of one millionth 10 -12
micro
nano n One thousandth of one millionth 10 -9
micro µ One millionth 10 -6
milli m One thousandth 10 -3
centi c One hundredth 10 -2
deci d One tenth 10 -1
kilo k One thousand 103
mega M One million 106
giga G One billion 109
tera T One trillion 1012
556
APPENDIX B
Answers to Exercises
CHAPTER 1
Al -1. Remember that in video tape recording, the signal recorded on the tape
is fm. The video signal is pre- emphasized in the modulator and de-
emphasized accordingly in the demodulator to improve the signal-to-
noise ratio. Ahead of the demodulator (where the signal is still
frequency-modulated rf), compensation is provided to allow for tape -
characteristic variations and head variations. Remember in this con-
nection that the highest video frequencies are in the lowest sidebands-
If the response at the higher fm frequencies is increased, lower video
frequencies are increased and highs decreased. If the response at the
lower fm frequencies is increased, lower video frequencies are de-
creased and highs increased.
A1-2. 50/5 = 10 -MHz bandwidth uncompensated. You will find that
modern solid -state video amplifiers, using good high -frequency tran-
sistors (very small capacitances) , use peaking circuits only in isolated
instances.
A1-3. Since
Velocity = Wavelength X Frequency,
then
Velocity
Wavelength =
Frequency
1560 = 156
microinches
10 MHz
A1-7. Since the standard video-to -sync ratio prevails, there is 0.45 volt sync
and 1.05 volts video. Since the modulation sensitivity is set such that
0.7 volt of video causes a deviation of 1.8 MHz, then for 1.05 volts:
0.7 1.05
1.8 x
SO
0.7x = 1.89
x = 2.7 MHz
Then
5 MHz + 2.7 MHz = 7.7 MHz for peak white
What happens here in most low -band recorders is that the shelf is
exceeded, and the modulator "goes to noise." This is a typical overload
condition in which there is a momentary dropout to "noise" for the
duration of the overload (on playback) .
The sync -tip frequency is found as follows:
0.3 - 0.45
0.7 x
so
0.3x = 0.315
x = 1.05 MHz
Then
5 MHz - 1.05 MHz = 3.95 MHz for sync tip
This problem has been solved with the old ratio- and -proportion
technique. A simple check is to reason that if the recorder modulation
sensitivity was set for 2.5 MHz total deviation when the input level
was 1 volt, then the sensitivity is 2.5 MHz /volt. So, if 1.5 volts of
signal is applied, the new deviation is:
(2.5) (1.5) = 3.75 MHz total deviation
This checks the above computations, since:
7.7 MHz peak white
- 3.95 MHz sync tip
3.75 MHz total deviation
CHAPTER 2
A2 -1. Sound track, video track, cue track, control track. NOTE: Modern tape
guides have a radius at the top so that video is no longer recorded on
the audio track at the top of the tape. In early machines, this trans-
verse video track was laid down across the audio track, then removed
by the following audio -track erase head prior to audio recording. The
audio erase head is, of course, still employed to remove any old c
spurious signal component.
A2 -2. No. The tape vacuum guide is moved.
A2 -3. Recording diameter = 0.004 + 2.064 = 2.068 in
ANSWERS TO EXERCISES 559
w.cuum guide.
A2 -8. 1590 ns. See the table in Appendix A.
212 -9. The active picture area is 63.5 µs minus blanking of 11 lts, or 52.5 µs.
So 10 inches corresponds to 52.5 µs, and one inch corresponds to
5.25 .ts. Therefore 1 l.ts is about 1/;-, inch, or between pia and 1/+ inch.
A2 -10. Review Section 2 -5 and Fig. 2 -13B. Departure at the center of the
tape is a measure of actual tip projection.
CHAPTER 3
interval of the signal being recorded (or local sync for local record-
ings).
A3 -3. The head- tachometer signal. This locks the capstan to the velocity
and phase of the headwheel (drum) operation.
A3 -4. The horizontal rate has a reference about every 33 lines. The vertical
reference is 60 Hz from the field rate of the recorded signal.
A3 -5. Same horizontal and vertical rates as in A3 -4. The initial field refer-
ence is usually from local sync, but no feedback information is given
to the capstan or headwheel for frequency and phase lock to local sync.
A3 -6. Same horizontal as above. Feedback information is given to the head
motor and capstan to lock the played -back tape sync to the local
vertical ( field) rate.
A3 -7. The horizontal rate is 15.75 kHz and the vertical rate is 60 Hz
(nominal values), locked in phase to local sync. This is "genlocked"
operation of the vtr.
A3 -8. (A) Line -to -line, or every 63.5 µs. (B) Eight to nine reference pulses
of 0.28 p.s corresponding to the 8 to 9 cycles of color -subcarrier burst,
once each line on the back porch of horizontal sync
A3 -9. Older machines and modern monochrome -only systems still use the
control -track and head -tachometer comparison of Fig. 3 -6. Later
systems use the basic method of Fig. 3 -8, with refinements. More
information on this is given in Chapter 7.
A3 -10. Velocity lock, then phase lock.
CHAPTER 4
A4 -1. (A) Master erase, audio erase, audio record /play, audio simulplay.
(B) Master erase, control -track record /play, cue erase, cue recor
play, control-track simulplay.
A4 -2. The position of the vacuum guide.
A4 -3. The capstan only. The supply and take -up reel motors merely provide
constant torque.
A4 -4. To take up excess tape if any speed variation occurs due to changes in
operating modes, and to stop the machine automatically if the tape
breaks or a reel runs out of tape.
A4 -5. To lift the tape from contact with the master erase head in either the
forward or reverse wind mode. This minimizes tape wear by eliminat-
ing contact during any operation not requiring contact with the mat«
erase head.
A4 -6. Air guides, tape lifter, and vacuum guide.
A4 -7. Headwheel air bearing and vacuum-guide air cylinder.
A4 -8. The one nearest to the +70 -volt trigger, in this case setup.
ANSWERS TO EXERCISES 561
CHAPTER 5
A5 -1. Only if the heads are within plus or minus 0.005 µs (5 ns) error.
This is within plus or minus 1.5 seconds of arc (approximately).
The latest high -band heads (also used for low band) meet this speci-
fication, and no electronic quadrature correctors are employed with
these heads. In practice, even when this tolerance is slightly exceeded,
built -in time -correction circuitry removes the error (Chapter 6) .
A5 -2. (A) 1 to 8 MHz
(B) 1to16MHz
A5 -3. (A) 4.28 MHz (H) 7.9 MHz
(B) 5.0 MHz (I) 10 MHz
(C) 6.8 MHz
(D) 5.5 MHz
(E) 5.79 MHz
(F) 6.5 MHz
(G) 7.06 MHz
A5 -4. Irregularities in tape coating that cause video loss, or "white flashes,"
for the duration of the dropout. This may result in tiny white dots or
much larger flashes, depending on the area of irregularity.
A5 -5. (A) Approximately 100 V (p -p) (B) Approximately 5 mV (p -p) .
.
CHAPTER 6
CHAPTER 7
A7 -1. Yes.
A7 -2. The control-track record current is increased until a just perceptible
inflection (shoulder) occurs in the region of the zero -axis crossing as
observed on the output of the simulplay control -track head.
A7 -3. Program audio head to assure that the audio signal is being recorded,
cue -track audio head to monitor the cue-track audio while recording,
and control-track head to assure that the control track is being laid
down and to permit proper adjustment of the record current.
A7 -4. You don't. This is determined by the arrangement shown in Fig. 7 -1.
A7 -5. By the TRACKING (CONTROL TRACK PHASE) knob, which adjusts
the delay in the control -track playback signal.
A7-6. Velocity lock followed by phase lock of the headwheel (drum) , from
which a trapezoid signal is formed. Then this trapezoid is sampled
with a pulse from:
1. Reference vertical from local sync.
2. Reference vertical from comparison of local vertical sync with
demodulated tape vertical sync.
3. Final step: Same as (2) but added "time- modulated" vertical
comparison to horizontal -sync errors.
A7 -7. (A) Reference frame pulse is compared in phase with pulse obtained
by dividing frequency of control-track signal by eight; reference frame
pulse is compared with tape frame pulse. (B) Vertical framing. (C)
Because the servos must be able to lock in to the closest horizontal
pulse. Therefore, the framing pulse must be dropped as a reference.
A7-8. (A) A 30 -pps signal recorded along with the control -track signal,
indicating occurrence of field 1. (B) Same as A. (C) A 30 -pps signal
derived from the vertical interval of the local sync generator. (D) A
30 -pps signal derived from the vertical interval of the demodulated
tape playback signal.
A7 -9. Yes. The tight control of time base requires timing (not necessarily
phasing) to the local sync generator.
A7 -10. To obtain greater sensitivity in microseconds per volt of error signal.
CHAPTER 8
A8-1. A reduction in rf amplitude, and therefore a video loss. "White
flashes" occur in the picture for the duration of the dropout. Drop-
ANSWERS TO EXERCISES 563
outs are caused by irregularities in the tape coating, either flaws in the
tape itself or dust and other foreign matter. They result in small white
dots or much larger flashes depending on the area of the irregularity.
A8 -2. The rf level from the recorded tape signal.
A8 -3. Not if the machine playing back to the dubbing system contains a
dropout compensation system. However, if it does not, any dub
(second -generation tape) will contain the same dropouts, and any
subsequent dub (third generation) cannot be replayed without the
dropouts even though a DOC is used. That is, the DOC acts only on
rf signal levels from the tape being played, and not on any "recorded -
in" dropout in the video signal.
A8 -4. The video from the previously scanned line is substituted. This is
possible because successive lines (or several lines) contain a large
amount of redundant information.
A8 -5. Since the substituted signal is delayed one line, the color phase is
shifted by 180 °. Therefore, luminance and chrominance information
are separated, and the chroma is phase- inverted to obtain the correct
color polarity.
CHAPTER 9
CHAPTER 10
A10 -1. (A) The new scene cannot be synchronous to the existing program
on the tape, and this causes servo upset, at both the start and stop of
the insert, until new lockup is achieved. (B) The distance between
the record and erase heads would cause a double (and nonsynchro-
nous) recording to be made for approximately 9 inches of tape.
410 -2. (A) The insert mode retains the already existing control -track signal
on the tape during the record mode. The new recording starts at the
desired place and ends before the old recording ends. Thus, it requires
an ingoing and an outgoing splice. (B) The add -on (assemble) mode
requires only an ingoing splice because it simply adds on to the
original recording. A new control track is recorded as in normal
recording. Synchronism with the original recording is maintained by
making the capstan servo a slave to a capstan memory circuit, which
has measured and stored the speed of the original recording.
410 -3. Improper phasing of the tachometer signal with respect to the origi-
nal recorded signal.
410 -4. (A) The original control -track signal. (B) A dc voltage from the
capstan memory circuit (controlling a vco).
410 -5. A computer system which enables the operator to "practice" a com-
plete editing procedure by watching a picture monitor (and listen-
ing to an audio monitor) without disturbing the original recording.
Then, when the entire procedure is satisfactory, the information is
stored again in the computer, and a single push button initiates auto-
matic editing.
CHAPTER 11
A11 -1. Start with midrange. The purpose of this control is to permit setting
the individual channel equalizers for equalization of all chroma (and
burst) levels. If any of the individual controls is nearing the limit
of its range, change the master equalizer accordingly.
A11 -2. In the latest systems, playback levels are controlled by agc. The con- t
trols marked MGC (manual gains) are effective only on test-signal
positions. Obviously, in an older tape system, you must first equalize
all levels (rf) as indicated on the cro at the switcher monitoring
position.
ANSWERS TO EXERCISES 565
A11 -11. By measuring the playback performance characteristic with the sine -
squared and window signals on the SMPTE standard alignment tape.
Then compare a local recording on the same playback. On modern
high -band systems, the performance should be better than that of the
monochrome standard alignment tape. The overall K- factor should
be well within 4 percent for the T -pulse and 2 percent for the 2T-
pulse.
A11 -12. From 1.8 MHz to 15 MHz (double -sideband rf signal) .
A11 -13. In the external -sync mode, local external reference sync from the
sync generator is added to the video signal. In the regenerated -sync
mode, the processing system forms new composite sync.
A11 -14. Yes. The RCA test tape is recorded on a 1.0331 -inch guide. Strictly
speaking, this would cause some difference between new and worn
heads. In practical everyday use, this is negligible, and you need have
no hesitation in setting the guide for any head (new or worn if more
than 1 mil of tip remains) against the standard tape.
A11 -15. 1 mil. Some users still employ a gauge to measure tip protrusion,
but
this is not advisable due to possible damage to a good head. The use
of the standard multiburst signal from a test tape (as described in
this chapter) is most satisfactory. The back -away method (to find the
point of departure) with precision measurement of the headwheel
radius is sometimes (but rarely) used.
A11 -16. No. When you record at a lower -than- standard penetration, tape
contact is lost as the head wears, and the result is about the same life
as with standard penetration. In addition, interchangeability is lost.
A11 -17. Here are the things that must be done: All sources to be lapped must
be synchronous, so the tape machine (s) must be operated in the
Pixlock (RCA) or automatic (Ampex) mode (tightest servo re-
quirement). Then each tape system must be phased horizontally for
the same timing at the leading edge of sync. Then the tape system
phase must be adjusted for proper burst phase at the mixing point.
It is also quite desirable that the blanking widths be approximately
the same, that the front -porch and back -porch positions be the same,
and that the positions of the two bursts on the back porch be very
close.
A11 -18. Some high -band tapes have a tendency to stick in the transport, and
this is normally the fault of the tape. High-band tapes have a very
smooth surface. A smooth surface lying against another smooth sur-
face has more of a tendency to cling to the guide, posts, etc., than a
rough surface against a smooth surface would have. This effect is
termed tape seizure or tape adhesion. The drag is actually so great
under this condition that capstan slippage occurs. If it happens often,
check all transport tensions, including the capstan pressure roller,
and thoroughly degauss the entire tape path.
ANSWERS TO EXERCISES 567
CHAPTER 12
Al2 -1. Axial alignment and axial position. Review Section 12 -2.
412 -2. "Tip ringing," a fault of the head itself, caused by deterioration of the
magnetic -gap material.
Al2 -3. Feed a multiburst signal to the system input. Response in the E -E
mode is immediately apparent by noting the difference between the
video output and the video input, with the machine in the stop
mode. Playback characteristics are judged by use of the standard test
tape. Recording characteristics then will be apparent by making a
recording of the local multiburst signal and playing this back on the
same and other machines of the common installation. The latter
step is simply to ascertain the degree of likeness or difference of all
machines.
Al2 -4. (A) 0.125H /cm X 25.
(B) 0.25H /cm X 25.
412 -5. The modulated 20T- pulse.
Al2 -6. Displacement of colors from the proper luminance position.
Al2 -7. Bent supply or take -up reels, improper tape- tension adjustments
(always refer to specific instruction book for system), dirty heads or
tape path, worn heads, contaminated capstan shaft or pinch roller,
insufficient capstan -to -tape pressure, bad tape.
Al2 -8. Magnetized head, dirty head, worn head, poor tape.
CHAPTER 13
A13 -1. Yes, if the tape is wound onto a conventional reel. The same type of
head and the same recording standards prevail for both the cassette/
cartridge and reel -to -reel systems.
A13 -2. Yes, if the tape length does not exceed the capacity of the cassette or
cartridge.
A13 -3. It can be purchased either way. If shared electronics is used, the other
machine must be an Ampex AVR -1.
A13 -4. It can be purchased either way. If shared electronics is used, the other
machine must be an RCA TR -60 or TR -70.
A13 -5. (A) Start of tape. (B) End of tape.
A13 -6. (A) A hole in the tape (sensed by a photocell) (B) A reflective
.
CHAPTER 14
A14 -1. (A) 30 seconds. (B) 60 seconds.
A14 -2. Four, the top and bottom surfaces of the top and bottom discs (two -
disc, four -head system).
A14 -3. The latest 30 seconds (60 seconds for alternate -field recording) is
maintained in storage. Prior recorded material is progressively
erased.
A14 -4. 60 revolutions per second (3600 rpm) .
A14 -5. 262% lines (one field), beginning and ending during the vertical -
blanking interval.
414-6. No. It is held stationary while the disc makes a complete revolution.
When the next head starts recording, the head that was recording
previously is stepped to a new position.
A14 -7. For conventional tape recording, the record currents are fed continu-
ously to all four rotating heads, and a separate head does the erasing.
In video disc recording, the record currents are gated on and off to
each head, and the head used for recording a track is also used to
erase that track by gating the erase current to the head at the proper
time.
A14 -8. 7 mils.
A14 -9. (A) 10 mils. (B) 3 mils.
A14 -10. Yes, continuously variable up to twice normal speed.
Index
Modulation-cont Phase-cont
index, 20 chroma, 493-494
process, basics of, 92 -95 differential, measurement of, 407 -408
Modulator inverter, chroma, 507 -509
afc, 111 -113 system, 361 -362
module, 498 Phasing, horizontal, 300
30-MHz, 503 Pinch roller, 78
Monitoring, built-in, 541 Pixlock, 206, 207 -212
Monochrome ATC, 158 -162 achieved circuit, 212
error gain, adjusting, 360 playback mode, 228 -229
Multiburst amplitude levels, 338 Platch circuit, 212 -213
Play
Near -lock, 432 control, 475
detector mode, 206
adjustments, 438 functions, servo, 194 -195
capstan servo, 435 -436 Playback
NEB, 158 alternate -field, 495 -496
Noise amplifier, 128-131
null adjustments, 444 -445 checks before, 380
positional, 162 equalization, setting, 357 -360
Nonphased color mode, 169 mode, 65 -69
Nonsynchronous signals, playback and Linelock, 229
recording of, 418 Pixlock, 228 -229
Normal Switchlock, 228
mode, 200 tonewheel, 228
speed, 491 -493 operation, 217 -223
Npc mode; see nonphased color mode synchronous, 69 -72
Numerical display panel, 545 -546 variable -speed, 541
Positional noise, 162
Open-loop operation, 273 -275 Preamplifier, video, 125 -128
Operation, system, 549 -555 Preplanning, 555
Optimizing Preset mode, 201
local sync generator, 396 -398 Processed video, 180 -182
video heads, 375 -379 Processing
Opu, 487 pulse, 177
Orientation of recorded signals, 27 -30 signal, 175 -182
Oscillator, slow- motion, 521 unit, output, 487
Oscilloscope method of measuring S/N video, 180
ratio, 409 -410 Processor module, dropout, 240 -245
Output Programmed electronic editing, princi-
processing unit, 487 ples of, 309 -314
video module, 507 Programmer, electronic editor, BE 210,
Overdeviation, 376 328 -331
Programming control circuits, sequen-
Pad, video, 400 -401 tial, 541
Panel Pulse (s )
automatic, control, 454 -456 clamp, 180
maintenance, 458 edit, 65, 214
manual control, 456 field identification, 522
memory, 549 frame, 214
numerical display, 545 -546 control- track, 65
record, 546 -549 generator, head-switch, 145
secondary control, 457 line sync, 522
setup, 458 memory -control, generation of,
signal control, 457 267 -271
Peaking circuits, 401 processing, 177
PEB, 158 servo reference, 522
Penetration error, 354 start, 177
Performance, overall, evaluation of, sync, afc, 179
391 -393 20T modulated, 405 -406
Phase vertical reference, 525
burst, 361 -362 Push -button logic, 516 -519
574 TELEVISION BROADCASTING